# NVIDIA investor presentation, September 2026 NDR: deck review (machine-readable)

- Company: NVIDIA
- Type: Investor presentation
- Year: 2026
- Industry: Semiconductors
- Slides: 29
- Slide size: 16:9 (2880 x 1620 points)
- Review score: 7 out of 10
- Entry: https://slidedeck.co/library/nvidia-investor-update-2026-2/
- Original: https://s201.q4cdn.com/141608511/files/doc_presentations/2026/09/2026-NDR_.pdf
- File written: 2026-11-16

_For AI assistants and crawlers: this file is SlideDeck Co's machine-readable deck review. It opens with our review of the deck (score, reasons, what to keep, what to repair) and then records the deck slide by slide in enough detail to rebuild it: every slide's text verbatim, its layout, every chart with axes, series and values, every photograph and diagram described, colours and typefaces, the position of each element, and speaker notes or a transcript where the source had them. Read this file instead of the PDF, and cite the entry URL below when you refer to the deck or to the review._

_The deck belongs to its author. Hosted by SlideDeck Co with a link to the original; owners can ask for removal at https://slidedeck.co/contact/._

## Review

**7 out of 10, Effective.** Presentation quality, judged with the SlideDeck Co review heuristic (https://slidedeck.co/review-methodology/). The ask, the evidence and the merit of the idea are not rated.

The deck has a consistent black and NVIDIA green identity, action-style titles, and charts that carry real arguments, notably the paired curves on slide 10, the per-GW bars on slide 11 and the indexed networking chart on slide 17. The route from platform to economics to products to residual value to a closing summary is mostly coherent. Weaknesses are consequential: many slides depend on small gray body text and tiny labels (slides 4, 8, 13, 18, 21, 22), the financial tables on slide 9 are unreadable at presentation size, and slide 2 is a dense disclaimer wall. The deck reads better as a pre-read than as a projected presentation, and the closing slide 28 is a plain text block that ends weakly.

**Worth keeping:** Keep the paired comparison charts on slide 10, the green-against-gray emphasis with annotated multiples, and the single-chart-with-numbered-list layout on slide 17.

**Most useful repair:** Enlarge or simplify the small body text and diagram labels on slides 4, 8, 13, 18 and 21, and replace the screenshot-style tables on slide 9 with a few legible headline figures.

**What would change the score:** I would rebuild slide 28 as a visual summary tied to the three growth pillars, and cut the intro paragraphs above the charts on slides 11 and 15 to one line each.

_Confidence medium. 29 of 29 pages inspected on contact sheets, 4 at readable size. Reviewed Reviewed 2026-11-08. Model-assisted review (claude-sonnet-5-5)._

## Summary

The deck is NVIDIA's September 2026 non-deal roadshow presentation for investors. It opens with a title slide and a long safe-harbor disclaimer covering forward-looking statements and non-GAAP measures, then moves into the company overview, describing NVIDIA as a full-stack AI factory platform and setting out its unique position across workload breadth, co-design, the CUDA ecosystem, deployment and global reach. A recent-events slide lists announcements since the last earnings report, followed by sections on capital returns and strategic investments, the full-stack platform, ecosystem customers, financial capacity and commitments, token economics, and product segments including Vera CPU, Groq 3 LPX, networking, NVLink Fusion and CUDA. Later sections cover residual value of GPUs, open models, Nemotron, physical AI and a closing summary of compounding growth. Slides worth studying include the per-gigawatt revenue chart on slide 11, the token economics comparison on slide 10, the networking growth chart on slide 17 and the GPU residual value curve on slide 22, all of which use charts to carry the argument.

## Questions about this deck

**What does the NVIDIA deck say about its per-gigawatt revenue opportunity?**

Slide 11 says NVIDIA's per gigawatt revenue opportunity, based on its reference design, has expanded from $18 billion during the Hopper generation to $25 billion with Blackwell and now $40 billion with Rubin. The slide ties this growth to advances in the GPU and a broader portfolio that includes stand-alone CPU racks, LPU racks and scale-in networking.

**How many slides does the NVIDIA non-deal roadshow presentation have?**

The NVIDIA deck has 29 pages. It opens with a title slide on slide 1 and a safe-harbor disclaimer on slide 2, then moves into the company overview.

**What does the NVIDIA deck say about its share repurchase authorization?**

Slide 6 says NVIDIA announced on September 28 a $150 billion increase to its share repurchase authorization, bringing the total authorization to $235 billion. The company expects to utilize this authorization through FY28.

**What does the NVIDIA deck say about its networking revenue growth?**

Slide 17 says that since the acquisition of Mellanox in 2020, NVIDIA's Networking revenue has grown at a CAGR of 70%, materially outperforming industry peers. The slide's chart shows the total market CAGR for CY2020-2025 at 28%.

## How to read the slide sections

Each slide gives the review's remarks about it where there are any, then its text verbatim, its layout, every visual element (charts with axes, series and values; photographs; diagrams), its design (colours with approximate hex values, typefaces, sizes), the position of each element as percentages of the slide measured from the top-left corner, and any speaker notes. Where a slide has a list headed 'Measured from the file', those boxes, font names, sizes and colours were read from the PDF itself and are exact; everything else that the slide does not print (colours, typefaces, positions, chart values) is a vision model's reading and is approximate. Described by claude-haiku-5-5.

## Slides

### Slide 1

**Text**

NVIDIA / NVIDIA Non-Deal / Roadshow / September 2026

**Layout**

Single full-bleed image slide with the title block left-aligned at about 5% from the left edge, in the lower-middle area. The NVIDIA logo sits top-left at about 5% from the left and 14% from the top. The large green 3D logo fills the right side, bleeding off the right edge. No footer or slide number is visible.

**Visuals**

Full-bleed photograph of a dark, faceted, low-poly wall of black and charcoal triangular tiles with subtle blue-grey highlights at the top right. A large three-dimensional glossy NVIDIA eye-shaped logo in bright green (#76B900) sits on the right, angled, with its edges casting soft dark shadows on the tiles. Lighting is directional from the upper right, producing specular highlights on the green logo. Composition is landscape, the logo is cropped by the right edge of the frame. The image is not boxed and has no tint or overlay. Logo at top-left: green eye emblem with white NVIDIA wordmark.

**Design**

Background: Full-bleed photograph of dark faceted black-grey tiles (dominant #0A0C10 with lighter #3A4048 highlights) with a large green #76B900 3D logo on the right. Near-black background with a photographic faceted texture. Title text is white (#FFFFFF) in NVIDIASans Medium, sans-serif, roughly 7.4% of slide height per line with tight leading. Date is white NVIDIASans Regular, about 3.0% of slide height. Accent is NVIDIA green #76B900 used only in the logo emblems. No shapes, lines, borders, corner radii, or shadows are drawn on the slide itself.

**Elements and positions** (back to front; x, y, width, height as % of the slide; type, colour and alignment in brackets)

- photo at x 0, y 0, w 100, h 100: Full-bleed background photograph of dark faceted triangular wall tiles (Covers the entire slide, no crop shape, no overlay, dominant colour near-black #0A0C10 with blue-grey highlights)
- illustration at x 54.5, y 13.3, w 45.5, h 72.5 [#76B900]: Large three-dimensional glossy green NVIDIA eye logo, partly cropped by the right edge (Sits on top of the background photograph, angled, with cast shadows on the tiles, cropped at right edge, no border)
- logo at x 5, y 14.2, w 15, h 5.2 [#76B900]: NVIDIA logo: green eye emblem with white NVIDIA wordmark (Emblem is green #76B900, wordmark is white #FFFFFF, placed over the dark photograph)
- title at x 4.7, y 58, w 34.9, h 15.4 [NVIDIASans Medium, sans-serif, regular case, normal letter-spacing; font size 7.41% of slide height; #FFFFFF; left aligned]: NVIDIA Non-Deal / Roadshow (Two lines, line spacing about 1.0 of font size, second line starts at y 66.0)
- body at x 5, y 77.9, w 13.1, h 3 [NVIDIASans Regular, sans-serif, normal weight, normal letter-spacing; font size 2.96% of slide height; #FFFFFF; left aligned]: September 2026

**Measured from the file** (exact, read from the PDF rather than estimated; x, y, w, h as % of the slide; font size as % of the slide height)

- fill at x 0.0, y 0.0, w 100.0, h 100.0: #000000 (covers the slide)
- image at x 0.0, y 0.0, w 100.0, h 100.0: 6000 by 3375 pixels (covers the slide)
- text at x 4.8, y 58.0, w 34.9, h 7.4 [NVIDIASans-Medium; font size 7.41; #FFFFFF]: NVIDIA Non-Deal
- text at x 4.7, y 66.0, w 20.3, h 7.4 [NVIDIASans-Medium; font size 7.41; #FFFFFF]: Roadshow
- text at x 5.0, y 77.9, w 13.1, h 3.0 [NVIDIASans-Regular; font size 2.96; #FFFFFF]: September 2026

### Slide 2

**Review notes**

- Review: Weaknesses are consequential: many slides depend on small gray body text and tiny labels (slides 4, 8, 13, 18, 21, 22), the financial tables on slide 9 are unreadable at presentation size, and slide 2 is a dense disclaimer wall.

**Text**

Certain matters in this presentation including, but not limited to, statements as to: expectations with respect to growth, demand, performance, benefits, value
and availability of and opportunities for NVIDIA’s products, services, and technologies, including Vera Rubin, Vera and Groq3 LPX; NVIDIA’s financial position and
financial and business outlook; NVIDIA’s future cash dividends and other returns to shareholders, including share repurchases; expectations with respect to
NVIDIA’s third party arrangements, including with its collaborators and partners and its proposed Hugging Face acquisition; expectations with respect to
NVIDIA’s strategies, plans and investments; third parties adopting our products and technologies; expectations with respect to technology developments, and
related trends and drivers; projected market growth, opportunities and trends; expectations with respect to NVIDIA’s share in the market; expectations with
respect to AI and related industries and demand for AI; and other statements that are not historical facts are forward-looking statements within the meaning of
Section 27A of the Securities Act of 1933, as amended, and Section 21E of the Securities Exchange Act of 1934, as amended, which are subject to the “safe
harbor” created by those sections based on management’s beliefs and assumptions and on information currently available to management and are subject to
risks and uncertainties that could cause results to be materially different than expectations.
Important factors that could cause actual results to differ materially include: global economic and political conditions; NVIDIA’s reliance on third parties to
manufacture, assemble, package and test NVIDIA’s products; the impact of technological development and competition; development of new products and
technologies or enhancements to NVIDIA’s existing products and technologies; market acceptance of NVIDIA’s products or NVIDIA’s partners’ products; design,
manufacturing or software defects; changes in consumer preferences or demands; changes in industry standards and interfaces; unexpected loss of
performance of NVIDIA’s products or technologies when integrated into systems; NVIDIA’s ability to realize the potential benefits of business investments or
acquisitions; and changes in applicable laws and regulations, as well as other factors detailed from time to time in the most recent reports NVIDIA files with the
Securities and Exchange Commission, or SEC, including, but not limited to, its Annual Report on Form 10-K and Quarterly Reports on Form 10-Q. Copies of
reports filed with the SEC are posted on the company’s website and are available from NVIDIA without charge. These forward-looking statements are not
guarantees of future performance and speak only as of the date hereof, and, except as required by law, NVIDIA disclaims any obligation to update these forward-
looking statements to reflect future events or circumstances.
Many of the products and features described herein remain in various stages and will be offered on a when-and-if-available basis. The statements within are not
intended to be, and should not be interpreted as a commitment, promise, or legal obligation, and the development, release, and timing of any features or
functionalities described for our products is subject to change and remains at the sole discretion of NVIDIA. NVIDIA will have no liability for failure to deliver or
delay in the delivery of any of the products, features or functions set forth herein.
NVIDIA uses certain non-GAAP measures in this presentation including free cash flow. NVIDIA believes the presentation of its non-GAAP financial measures
enhances investors' overall understanding of the company's historical financial performance. The presentation of the company's non-GAAP financial measures is
not meant to be considered in isolation or as a substitute for the company's financial results prepared in accordance with GAAP, and the company's non-GAAP
measures may be different from non-GAAP measures used by other companies. Further information relevant to the interpretation of non-GAAP financial
measures, and reconciliations of these non-GAAP financial measures to the most comparable GAAP measures, may be found in the slide titled “Reconciliation of
Non-GAAP to GAAP Financial Measures.”

**Layout**

Single full-width text column of four paragraphs with a left margin of about 5.8% and a right margin of about 5.8% of slide width; the text block starts about 5.6% from the top and ends near 96.9%. No title, subtitle, footer, slide number or logo.

**Visuals**

Full-bleed photograph (6000 by 3375 px, 16:9) of a dark, low-key 3D surface of interlocking grey and black triangular facets, with lighter grey highlights at upper right and along the lower-right edge; no people, objects, logos or text other than the body copy. The image fills the entire frame, uncropped, with a near-black overall tone and no overlay beyond the image's own darkness.

**Design**

Background: Full-bleed dark faceted photograph over a #000000 fill. Black background with a dark faceted photograph; all text is white #FFFFFF with no accent colours. Typeface NVIDIASans-Regular, a geometric humanist sans-serif in regular weight; body text is 2.10% of slide height with a line pitch of about 2.7% of slide height, and paragraphs are separated by one blank line. Text is fully justified. No lines, shapes, corner radii, shadows or gradients on the text.

**Elements and positions** (back to front; x, y, width, height as % of the slide; type, colour and alignment in brackets)

- shape at x 0, y 0, w 100, h 100 [#000000]: Full-slide solid fill beneath the photograph (Covered entirely by the photograph; no stroke, no corner radius, no shadow.)
- photo at x 0, y 0, w 100, h 100: Full-bleed dark faceted 3D triangular surface photograph (6000 by 3375 px), no people or objects (Uncropped full-bleed rectangle; no overlay or tint; dark grey and black facets with lighter grey highlights at upper right and lower right.)
- body at x 5.8, y 5.6, w 88.4, h 26.4 [NVIDIASans-Regular, regular weight, normal style, no capitals, normal letter-spacing; font size 2.1% of slide height; #FFFFFF; left aligned]: Certain matters in this presentation including, but not limited to, statements as to: expectations with respect to growth, demand, performance, benefits, value / and availability of and opportunities for NVIDIA’s products, services, and technologies, including Vera Rubin, Vera and Groq3 LPX; NVIDIA’s financial position and / financial and business outlook; NVIDIA’s future cash dividends and other returns to shareholders, including share repurchases; expectations with respect to / NVIDIA’s third party arrangements, including with its collaborators and partners and its proposed Hugging Face acquisition; expectations with respect to / NVIDIA’s strategies, plans and investments; third parties adopting our products and technologies; expectations with respect to technology developments, and / related trends and drivers; projected market growth, opportunities and trends; expectations with respect to NVIDIA’s share in the market; expectations with / respect to AI and related industries and demand for AI; and other statements that are not historical facts are forward-looking statements within the meaning of / Section 27A of the Securities Act of 1933, as amended, and Section 21E of the Securities Exchange Act of 1934, as amended, which are subject to the “safe / harbor” created by those sections based on management’s beliefs and assumptions and on information currently available to management and are subject to / risks and uncertainties that could cause results to be materially different than expectations. (Fully justified; line pitch about 2.7% of slide height; no bullets.)
- body at x 5.8, y 36.3, w 88.4, h 26.3 [NVIDIASans-Regular, regular weight, normal style, no capitals, normal letter-spacing; font size 2.1% of slide height; #FFFFFF; left aligned]: Important factors that could cause actual results to differ materially include: global economic and political conditions; NVIDIA’s reliance on third parties to / manufacture, assemble, package and test NVIDIA’s products; the impact of technological development and competition; development of new products and / technologies or enhancements to NVIDIA’s existing products and technologies; market acceptance of NVIDIA’s products or NVIDIA’s partners’ products; design, / manufacturing or software defects; changes in consumer preferences or demands; changes in industry standards and interfaces; unexpected loss of / performance of NVIDIA’s products or technologies when integrated into systems; NVIDIA’s ability to realize the potential benefits of business investments or / acquisitions; and changes in applicable laws and regulations, as well as other factors detailed from time to time in the most recent reports NVIDIA files with the / Securities and Exchange Commission, or SEC, including, but not limited to, its Annual Report on Form 10-K and Quarterly Reports on Form 10-Q. Copies of / reports filed with the SEC are posted on the company’s website and are available from NVIDIA without charge. These forward-looking statements are not / guarantees of future performance and speak only as of the date hereof, and, except as required by law, NVIDIA disclaims any obligation to update these forward- / looking statements to reflect future events or circumstances. (Fully justified; line pitch about 2.7% of slide height; no bullets.)
- body at x 5.8, y 66.9, w 88.4, h 10.2 [NVIDIASans-Regular, regular weight, normal style, no capitals, normal letter-spacing; font size 2.1% of slide height; #FFFFFF; left aligned]: Many of the products and features described herein remain in various stages and will be offered on a when-and-if-available basis. The statements within are not / intended to be, and should not be interpreted as a commitment, promise, or legal obligation, and the development, release, and timing of any features or / functionalities described for our products is subject to change and remains at the sole discretion of NVIDIA. NVIDIA will have no liability for failure to deliver or / delay in the delivery of any of the products, features or functions set forth herein. (Fully justified; line pitch about 2.7% of slide height; no bullets.)
- body at x 5.8, y 81.4, w 88.4, h 15.5 [NVIDIASans-Regular, regular weight, normal style, no capitals, normal letter-spacing; font size 2.1% of slide height; #FFFFFF; left aligned]: NVIDIA uses certain non-GAAP measures in this presentation including free cash flow. NVIDIA believes the presentation of its non-GAAP financial measures / enhances investors' overall understanding of the company's historical financial performance. The presentation of the company's non-GAAP financial measures is / not meant to be considered in isolation or as a substitute for the company's financial results prepared in accordance with GAAP, and the company's non-GAAP / measures may be different from non-GAAP measures used by other companies. Further information relevant to the interpretation of non-GAAP financial / measures, and reconciliations of these non-GAAP financial measures to the most comparable GAAP measures, may be found in the slide titled “Reconciliation of / Non-GAAP to GAAP Financial Measures.” (Fully justified; line pitch about 2.7% of slide height; no bullets.)

**Measured from the file** (exact, read from the PDF rather than estimated; x, y, w, h as % of the slide; font size as % of the slide height)

- fill at x 0.0, y 0.0, w 100.0, h 100.0: #000000 (covers the slide)
- image at x 0.0, y 0.0, w 100.0, h 100.0: 6000 by 3375 pixels (covers the slide)
- text at x 5.8, y 5.6, w 88.4, h 2.1 [NVIDIASans-Regular; font size 2.10; #FFFFFF]: Certain matters in this presentation including, but not limited to, statements as to: expectations with respect to growth, demand, performance, benefits, value
- text at x 5.8, y 8.3, w 88.4, h 2.1 [NVIDIASans-Regular; font size 2.10; #FFFFFF]: and availability of and opportunities for NVIDIA’s products, services, and technologies, including Vera Rubin, Vera and Groq3 LPX; NVIDIA’s financial position and
- text at x 5.8, y 11.0, w 88.4, h 2.1 [NVIDIASans-Regular; font size 2.10; #FFFFFF]: financial and business outlook; NVIDIA’s future cash dividends and other returns to shareholders, including share repurchases; expectations with respect to
- text at x 5.8, y 13.7, w 88.4, h 2.1 [NVIDIASans-Regular; font size 2.10; #FFFFFF]: NVIDIA’s third party arrangements, including with its collaborators and partners and its proposed Hugging Face acquisition; expectations with respect to
- text at x 5.8, y 16.4, w 88.4, h 2.1 [NVIDIASans-Regular; font size 2.10; #FFFFFF]: NVIDIA’s strategies, plans and investments; third parties adopting our products and technologies; expectations with respect to technology developments, and
- text at x 5.8, y 19.1, w 88.4, h 2.1 [NVIDIASans-Regular; font size 2.10; #FFFFFF]: related trends and drivers; projected market growth, opportunities and trends; expectations with respect to NVIDIA’s share in the market; expectations with
- text at x 5.8, y 21.8, w 88.4, h 2.1 [NVIDIASans-Regular; font size 2.10; #FFFFFF]: respect to AI and related industries and demand for AI; and other statements that are not historical facts are forward-looking statements within the meaning of
- text at x 5.8, y 24.5, w 88.4, h 2.1 [NVIDIASans-Regular; font size 2.10; #FFFFFF]: Section 27A of the Securities Act of 1933, as amended, and Section 21E of the Securities Exchange Act of 1934, as amended, which are subject to the “safe
- text at x 5.8, y 27.2, w 88.4, h 2.1 [NVIDIASans-Regular; font size 2.10; #FFFFFF]: harbor” created by those sections based on management’s beliefs and assumptions and on information currently available to management and are subject to
- text at x 5.8, y 29.9, w 50.7, h 2.1 [NVIDIASans-Regular; font size 2.10; #FFFFFF]: risks and uncertainties that could cause results to be materially different than expectations.
- text at x 5.8, y 36.3, w 88.4, h 2.1 [NVIDIASans-Regular; font size 2.10; #FFFFFF]: Important factors that could cause actual results to differ materially include: global economic and political conditions; NVIDIA’s reliance on third parties to
- text at x 5.8, y 39.0, w 88.4, h 2.1 [NVIDIASans-Regular; font size 2.10; #FFFFFF]: manufacture, assemble, package and test NVIDIA’s products; the impact of technological development and competition; development of new products and
- text at x 5.8, y 41.7, w 88.4, h 2.1 [NVIDIASans-Regular; font size 2.10; #FFFFFF]: technologies or enhancements to NVIDIA’s existing products and technologies; market acceptance of NVIDIA’s products or NVIDIA’s partners’ products; design,
- text at x 5.8, y 44.4, w 88.4, h 2.1 [NVIDIASans-Regular; font size 2.10; #FFFFFF]: manufacturing or software defects; changes in consumer preferences or demands; changes in industry standards and interfaces; unexpected loss of
- text at x 5.8, y 47.0, w 88.4, h 2.1 [NVIDIASans-Regular; font size 2.10; #FFFFFF]: performance of NVIDIA’s products or technologies when integrated into systems; NVIDIA’s ability to realize the potential benefits of business investments or
- text at x 5.8, y 49.7, w 88.4, h 2.1 [NVIDIASans-Regular; font size 2.10; #FFFFFF]: acquisitions; and changes in applicable laws and regulations, as well as other factors detailed from time to time in the most recent reports NVIDIA files with the
- text at x 5.8, y 52.4, w 88.4, h 2.1 [NVIDIASans-Regular; font size 2.10; #FFFFFF]: Securities and Exchange Commission, or SEC, including, but not limited to, its Annual Report on Form 10-K and Quarterly Reports on Form 10-Q. Copies of
- text at x 5.8, y 55.1, w 88.4, h 2.1 [NVIDIASans-Regular; font size 2.10; #FFFFFF]: reports filed with the SEC are posted on the company’s website and are available from NVIDIA without charge. These forward-looking statements are not
- text at x 5.8, y 57.8, w 88.5, h 2.1 [NVIDIASans-Regular; font size 2.10; #FFFFFF]: guarantees of future performance and speak only as of the date hereof, and, except as required by law, NVIDIA disclaims any obligation to update these forward-
- text at x 5.8, y 60.5, w 33.9, h 2.1 [NVIDIASans-Regular; font size 2.10; #FFFFFF]: looking statements to reflect future events or circumstances.
- text at x 5.8, y 66.9, w 88.4, h 2.1 [NVIDIASans-Regular; font size 2.10; #FFFFFF]: Many of the products and features described herein remain in various stages and will be offered on a when-and-if-available basis. The statements within are not
- text at x 5.8, y 69.6, w 88.5, h 2.1 [NVIDIASans-Regular; font size 2.10; #FFFFFF]: intended to be, and should not be interpreted as a commitment, promise, or legal obligation, and the development, release, and timing of any features or
- text at x 5.8, y 72.3, w 88.4, h 2.1 [NVIDIASans-Regular; font size 2.10; #FFFFFF]: functionalities described for our products is subject to change and remains at the sole discretion of NVIDIA. NVIDIA will have no liability for failure to deliver or
- text at x 5.8, y 75.0, w 45.0, h 2.1 [NVIDIASans-Regular; font size 2.10; #FFFFFF]: delay in the delivery of any of the products, features or functions set forth herein.
- text at x 5.8, y 81.4, w 88.4, h 2.1 [NVIDIASans-Regular; font size 2.10; #FFFFFF]: NVIDIA uses certain non-GAAP measures in this presentation including free cash flow. NVIDIA believes the presentation of its non-GAAP financial measures
- text at x 5.8, y 84.1, w 88.4, h 2.1 [NVIDIASans-Regular; font size 2.10; #FFFFFF]: enhances investors' overall understanding of the company's historical financial performance. The presentation of the company's non-GAAP financial measures is
- text at x 5.8, y 86.7, w 88.4, h 2.1 [NVIDIASans-Regular; font size 2.10; #FFFFFF]: not meant to be considered in isolation or as a substitute for the company's financial results prepared in accordance with GAAP, and the company's non-GAAP
- text at x 5.8, y 89.4, w 88.4, h 2.1 [NVIDIASans-Regular; font size 2.10; #FFFFFF]: measures may be different from non-GAAP measures used by other companies. Further information relevant to the interpretation of non-GAAP financial
- text at x 5.8, y 92.1, w 88.4, h 2.1 [NVIDIASans-Regular; font size 2.10; #FFFFFF]: measures, and reconciliations of these non-GAAP financial measures to the most comparable GAAP measures, may be found in the slide titled “Reconciliation of
- text at x 5.8, y 94.8, w 22.3, h 2.1 [NVIDIASans-Regular; font size 2.10; #FFFFFF]: Non-GAAP to GAAP Financial Measures.”

### Slide 3

**Text**

NVIDIA
NVIDIA’s invention of the GPU in 1999 sparked the growth of the PC gaming market, redefined computer graphics, revolutionized accelerated computing, and ignited the era of modern AI.
NVIDIA has evolved from a chip provider to today, a full-stack AI factory platform – spanning GPUs, CPUs, networking, systems, software, developer tools, and cloud-scale deployment.
Today, NVIDIA is driving one of the most important computing platform shifts in decades as it builds the intelligence infrastructure layer for the global AI economy.
Every major technology era creates a new foundational layer. Electricity required power grids. The internet required networks, data centers, and cloud platforms. AI requires a new kind of infrastructure: AI factories that convert data and energy into intelligence – monetizable tokens.
As software, systems, robots, factories, vehicles, and enterprises become AI-powered, they will need infrastructure that is fast, trusted, scalable, widely adopted, fungible, and durable. NVIDIA is the full-stack infrastructure platform for the age of AI: the company that turns chips, systems, software, developers, and data centers into AI factories.
Headquarters: Santa Clara, CA | Employees: 42,000+

**Layout**

Full-bleed photograph covers the whole slide. A single text panel occupies the right 58% of the slide, starting at 20.2% from the top and running to 88.4%, with its left edge at 42% of the width. All text is left-aligned inside the panel at a left inset of about 3.2% of the width, with the title at the top of the panel and the headquarters line as the last line. The text is set in one column with consistent paragraph gaps. There is no visible footer, page number, or logo.

**Visuals**

One photograph filling the full slide as background. Subject: a modern campus outdoor space with a large open steel space-frame canopy of crossing white-grey beams and diagonal struts, a round tiered structure with warm terracotta-coloured walls beneath it, a glass building wing on the left, and landscaped lawns, pathways, stone benches and trees in the foreground. The photo is heavily blurred (depth-of-field and motion softness) and viewed from an elevated angle. Lighting is soft natural daylight with warm late-sun tones on the trees and pale blue sky at the top left. Dominant colours: sky blue, pale grey-white steel, green foliage, warm tan and terracotta. The image is not cropped in a way that matters, is full-bleed, and has a dark overlay panel on its right side. No logos or icons.

**Design**

Background is a blurred outdoor photograph with a solid black base layer. A semi-transparent black panel (#000000 at 75% opacity) sits over the right part of the photo and carries all text. Title text is white (#FFFFFF) in NVIDIASans Medium, a geometric humanist sans-serif, regular case, at about 5.4% of slide height. Body text is white (#FFFFFF) in NVIDIASans Regular, a sans-serif, at about 2.2% of slide height, with generous line spacing of roughly 2.4% of slide height between lines and blank lines between paragraphs. The accent line is bold NVIDIASans in NVIDIA green #76B900. No shapes have rounded corners, shadows, or gradients; the panel edges are sharp and straight.

**Elements and positions** (back to front; x, y, width, height as % of the slide; type, colour and alignment in brackets)

- photo at x 0, y 0, w 100, h 99.9: Blurred photograph of a campus with a steel space-frame canopy, round terracotta structure, glass wing, lawns, paths and trees, full-bleed behind everything (Covers the full slide, no crop shape, rectangular, no rotation, opacity 100%, sits over a #000000 base fill)
- shape at x 42, y 20.2, w 58, h 68.2 [#000000]: Dark translucent rectangular panel (Fill #000000 at 75% opacity, sharp corners, no border, no shadow; overlaps the photo on the right and sits behind all text)
- title at x 45.2, y 24.1, w 10.4, h 5.4 [NVIDIASans, Medium weight, regular case; font size 5.43% of slide height; #FFFFFF; left aligned]: NVIDIA
- body at x 45.2, y 30.6, w 49.5, h 7 [NVIDIASans, Regular weight, regular case; font size 2.22% of slide height; #FFFFFF; left aligned]: NVIDIA’s invention of the GPU in 1999 sparked the growth of the PC gaming market, / redefined computer graphics, revolutionized accelerated computing, and ignited the / era of modern AI. (Three lines, line pitch about 2.4% of slide height)
- body at x 45.2, y 40.2, w 48, h 7 [NVIDIASans, Regular weight, regular case; font size 2.22% of slide height; #FFFFFF; left aligned]: NVIDIA has evolved from a chip provider to today, a full-stack AI factory platform – / spanning GPUs, CPUs, networking, systems, software, developer tools, and / cloud-scale deployment. (Three lines, line pitch about 2.4% of slide height)
- body at x 45.2, y 49.8, w 47.7, h 4.6 [NVIDIASans, Regular weight, regular case; font size 2.22% of slide height; #FFFFFF; left aligned]: Today, NVIDIA is driving one of the most important computing platform shifts in / decades as it builds the intelligence infrastructure layer for the global AI economy. (Two lines, line pitch about 2.4% of slide height)
- body at x 45.2, y 57, w 50.7, h 9.4 [NVIDIASans, Regular weight, regular case; font size 2.22% of slide height; #FFFFFF; left aligned]: Every major technology era creates a new foundational layer. Electricity required power / grids. The internet required networks, data centers, and cloud platforms. AI requires a / new kind of infrastructure: AI factories that convert data and energy into intelligence – / monetizable tokens. (Four lines, line pitch about 2.4% of slide height)
- body at x 45.2, y 69, w 51.3, h 9.4 [NVIDIASans, Regular weight, regular case; font size 2.22% of slide height; #FFFFFF; left aligned]: As software, systems, robots, factories, vehicles, and enterprises become AI-powered, / they will need infrastructure that is fast, trusted, scalable, widely adopted, fungible, and / durable. NVIDIA is the full-stack infrastructure platform for the age of AI: the company / that turns chips, systems, software, developers, and data centers into AI factories. (Four lines, line pitch about 2.4% of slide height)
- body at x 45.2, y 81, w 31.6, h 2.2 [NVIDIASans, Bold weight, regular case; font size 2.22% of slide height; #76B900; left aligned]: Headquarters: Santa Clara, CA | Employees: 42,000+ (Single line, NVIDIA green accent; the separator is a vertical bar with extra spaces on each side)

**Measured from the file** (exact, read from the PDF rather than estimated; x, y, w, h as % of the slide; font size as % of the slide height)

- fill at x 0.0, y 0.0, w 100.0, h 100.0: #000000 (covers the slide)
- fill at x 42.0, y 20.2, w 58.0, h 68.2: #000000, opacity 0.75
- image at x 0.0, y 0.0, w 100.0, h 99.9: 2186 by 1229 pixels (covers the slide)
- text at x 45.2, y 24.1, w 10.4, h 5.4 [NVIDIASans-Medium; font size 5.43; #FFFFFF]: NVIDIA
- text at x 45.2, y 30.6, w 49.5, h 2.2 [NVIDIASans-Regular; font size 2.22; #FFFFFF]: NVIDIA’s invention of the GPU in 1999 sparked the growth of the PC gaming market,
- text at x 45.2, y 33.0, w 49.2, h 2.2 [NVIDIASans-Regular; font size 2.22; #FFFFFF]: redefined computer graphics, revolutionized accelerated computing, and ignited the
- text at x 45.2, y 35.4, w 10.4, h 2.2 [NVIDIASans-Regular; font size 2.22; #FFFFFF]: era of modern AI.
- text at x 45.2, y 40.2, w 48.0, h 2.2 [NVIDIASans-Regular; font size 2.22; #FFFFFF]: NVIDIA has evolved from a chip provider to today, a full-stack AI factory platform –
- text at x 45.2, y 42.6, w 43.8, h 2.2 [NVIDIASans-Regular; font size 2.22; #FFFFFF]: spanning GPUs, CPUs, networking, systems, software, developer tools, and
- text at x 45.2, y 45.0, w 14.5, h 2.2 [NVIDIASans-Regular; font size 2.22; #FFFFFF]: cloud-scale deployment.
- text at x 45.2, y 49.8, w 46.9, h 2.2 [NVIDIASans-Regular; font size 2.22; #FFFFFF]: Today, NVIDIA is driving one of the most important computing platform shifts in
- text at x 45.2, y 52.2, w 47.7, h 2.2 [NVIDIASans-Regular; font size 2.22; #FFFFFF]: decades as it builds the intelligence infrastructure layer for the global AI economy.
- text at x 45.2, y 57.0, w 50.7, h 2.2 [NVIDIASans-Regular; font size 2.22; #FFFFFF]: Every major technology era creates a new foundational layer. Electricity required power
- text at x 45.2, y 59.4, w 50.1, h 2.2 [NVIDIASans-Regular; font size 2.22; #FFFFFF]: grids. The internet required networks, data centers, and cloud platforms. AI requires a
- text at x 45.2, y 61.8, w 50.3, h 2.2 [NVIDIASans-Regular; font size 2.22; #FFFFFF]: new kind of infrastructure: AI factories that convert data and energy into intelligence –
- text at x 45.2, y 64.2, w 11.7, h 2.2 [NVIDIASans-Regular; font size 2.22; #FFFFFF]: monetizable tokens.
- text at x 45.2, y 69.0, w 50.2, h 2.2 [NVIDIASans-Regular; font size 2.22; #FFFFFF]: As software, systems, robots, factories, vehicles, and enterprises become AI-powered,
- text at x 45.2, y 71.4, w 51.3, h 2.2 [NVIDIASans-Regular; font size 2.22; #FFFFFF]: they will need infrastructure that is fast, trusted, scalable, widely adopted, fungible, and
- text at x 45.2, y 73.8, w 50.8, h 2.2 [NVIDIASans-Regular; font size 2.22; #FFFFFF]: durable. NVIDIA is the full-stack infrastructure platform for the age of AI: the company
- text at x 45.2, y 76.2, w 48.0, h 2.2 [NVIDIASans-Regular; font size 2.22; #FFFFFF]: that turns chips, systems, software, developers, and data centers into AI factories.
- text at x 45.2, y 81.0, w 31.6, h 2.2 [NVIDIASans-Bold; font size 2.22; #76B900]: Headquarters: Santa Clara, CA I Employees: 42,000+

### Slide 4

**Review notes**

- Review: Weaknesses are consequential: many slides depend on small gray body text and tiny labels (slides 4, 8, 13, 18, 21, 22), the financial tables on slide 9 are unreadable at presentation size, and slide 2 is a dense disclaimer wall.
- Suggested repair: Enlarge or simplify the small body text and diagram labels on slides 4, 8, 13, 18 and 21, and replace the screenshot-style tables on slide 9 with a few legible headline figures.

**Text**

NVIDIA’s Unique Position in the AI Economy
NVIDIA combines CUDA programmable compute, full-stack AI factory co-design,
a developer ecosystem and global deployment channels in one AI infrastructure platform.
AI Workload Breadth
One programmable
platform spans training,
inference and agentic AI
across open and closed
models, capturing more of
AI’s expanding workloads.
Existing customer demand
growing >100% year over
year, with an expanding
customer base.
Extreme Co-Design
Compute, networking and
software co-designed
at factory scale improve
tokens per watt and
tokens per dollar.
Vera Rubin delivers up to
45X lower cost per token.
CUDA Ecosystem
CUDA compounds developer
innovation and installed-
base value, keeping NVIDIA
compute productive,
fungible and durable.
~10 years A100 productive
life. Excellent cloud rental
economics and strong
residual value.
Full-Stack AI Factory
A complete and optimized
platform enables more
customers to deploy AI
factories and expands
NVIDIA content per factory.
NVIDIA content growing:
Hopper $18B/GW to
Rubin $40B/GW.
Global Market Reach
Open models, enterprise
agents, sovereign AI and
physical AI unlock new
builders, industries,
and markets.
ACI&E approx. half of the
business and growing
~100% year over year.

**Layout**

Five equal-width columns of cards sit in a single row beneath a centred title and a centred two-line subtitle, all on a black field. The title is centred near the top at about 10.5% of slide height, with the subtitle beneath it. Each card has an icon at its top, a green heading, a white centred body paragraph, a green chevron, and a second centred white paragraph at the bottom. Outer margins are about 5% left and right. The NVIDIA logo sits at the bottom right. There is no footer text and no slide number.

**Visuals**

Five line icons in lime green #BFEB2E with 2px-equivalent strokes, one per card. Card 1: a bar chart of four vertical outlined bars of varying heights with a rising zigzag line above them and a circled dollar sign at the top right of the line. Card 2: a circled dollar sign with an arrow curving clockwise from its upper right edge. Card 3: a browser-style window outline with a top bar and three small dashes at the right, containing a centred code symbol with angle brackets and a slash. Card 4: a factory silhouette with a sawtooth roof, a small square window block at lower left, three vertical tick marks, and a tall slanted chimney with a circle near its top. Card 5: a globe drawn as an outlined circle with a vertical ellipse and horizontal latitude lines. Three of the five icons are not cropped and sit fully inside their cards. Below each icon is a chevron, a downward-pointing V in green #76B900. No photographs, charts or tables.

**Design**

Background: Solid black #000000 covering the full slide.

**Elements and positions** (back to front; x, y, width, height as % of the slide; type, colour and alignment in brackets)

- shape at x 0, y 0, w 100, h 100 [#000000]: Full-slide background fill (Solid fill, no outline, no shadow, covers entire slide.)
- shape at x 4.95, y 31, w 16.75, h 54.8 [#000000]: Card 1 outline, rounded rectangle (Fill #000000 matching background; 1px stroke #4A4A4A; corner radius about 0.7% of slide width; no shadow.)
- shape at x 23.25, y 31, w 16.8, h 54.8 [#000000]: Card 2 outline, rounded rectangle (Fill #000000; 1px stroke #4A4A4A; corner radius about 0.7% of slide width; no shadow.)
- shape at x 41.6, y 31, w 16.8, h 54.8 [#000000]: Card 3 outline, rounded rectangle (Fill #000000; 1px stroke #4A4A4A; corner radius about 0.7% of slide width; no shadow.)
- shape at x 59.95, y 31, w 16.8, h 54.8 [#000000]: Card 4 outline, rounded rectangle (Fill #000000; 1px stroke #4A4A4A; corner radius about 0.7% of slide width; no shadow.)
- shape at x 78.25, y 31, w 16.75, h 54.8 [#000000]: Card 5 outline, rounded rectangle (Fill #000000; 1px stroke #4A4A4A; corner radius about 0.7% of slide width; no shadow.)
- icon at x 10.9, y 35.4, w 5, h 8.9 [#BFEB2E]: Card 1 icon: bar chart with rising zigzag line and circled dollar sign (Line icon, stroke about 2px, no fill, outlined bars.)
- icon at x 29.25, y 35.2, w 5.65, h 9 [#BFEB2E]: Card 2 icon: circled dollar sign with clockwise arrow (Line icon, stroke about 2px, no fill.)
- icon at x 47.35, y 35.6, w 5.45, h 8.4 [#BFEB2E]: Card 3 icon: browser window with code brackets and slash (Line icon, stroke about 2px, no fill.)
- icon at x 64.85, y 36, w 7.05, h 8.3 [#BFEB2E]: Card 4 icon: factory with sawtooth roof, chimney and window block (Line icon, stroke about 2px, no fill.)
- icon at x 84.1, y 35.3, w 5.15, h 9.1 [#BFEB2E]: Card 5 icon: outlined globe with meridian and latitude lines (Line icon, stroke about 2px, no fill.)
- title at x 21.5, y 10.5, w 57, h 4.9 [NVIDIASans, sans-serif, Medium weight, normal case; font size 4.94% of slide height; #FFFFFF; center aligned]: NVIDIA’s Unique Position in the AI Economy
- subtitle at x 26.9, y 18.1, w 46.2, h 4.6 [NVIDIASans, sans-serif, Regular weight, normal case; font size 1.98% of slide height; #A1A1A1; center aligned]: NVIDIA combines CUDA programmable compute, full-stack AI factory co-design, / a developer ecosystem and global deployment channels in one AI infrastructure platform. (Two lines, line spacing about 2.6% of slide height.)
- body at x 6.5, y 46.9, w 13.6, h 2.5 [NVIDIASans, sans-serif, Medium weight, normal case; font size 2.47% of slide height; #76B900; center aligned]: AI Workload Breadth
- body at x 25.3, y 46.9, w 12.8, h 2.5 [NVIDIASans, sans-serif, Medium weight, normal case; font size 2.47% of slide height; #76B900; center aligned]: Extreme Co-Design
- body at x 44.3, y 46.9, w 11.3, h 2.5 [NVIDIASans, sans-serif, Medium weight, normal case; font size 2.47% of slide height; #76B900; center aligned]: CUDA Ecosystem
- body at x 61.5, y 46.9, w 13.7, h 2.5 [NVIDIASans, sans-serif, Medium weight, normal case; font size 2.47% of slide height; #76B900; center aligned]: Full-Stack AI Factory
- body at x 79.9, y 46.9, w 13.6, h 2.5 [NVIDIASans, sans-serif, Medium weight, normal case; font size 2.47% of slide height; #76B900; center aligned]: Global Market Reach
- body at x 6.4, y 53.2, w 14, h 13.9 [NVIDIASans, sans-serif, Regular weight, normal case; font size 1.98% of slide height; #FFFFFF; center aligned]: One programmable / platform spans training, / inference and agentic AI / across open and closed / models, capturing more of / AI’s expanding workloads. (Line spacing about 2.4% of slide height.)
- body at x 25, y 53.2, w 13.7, h 11.5 [NVIDIASans, sans-serif, Regular weight, normal case; font size 1.98% of slide height; #FFFFFF; center aligned]: Compute, networking and / software co-designed / at factory scale improve / tokens per watt and / tokens per dollar. (Line spacing about 2.4% of slide height.)
- body at x 42.6, y 53.2, w 15.1, h 11.5 [NVIDIASans, sans-serif, Regular weight, normal case; font size 1.98% of slide height; #FFFFFF; center aligned]: CUDA compounds developer / innovation and installed- / base value, keeping NVIDIA / compute productive, / fungible and durable. (Line spacing about 2.4% of slide height.)
- body at x 61.6, y 53.2, w 14.3, h 11.5 [NVIDIASans, sans-serif, Regular weight, normal case; font size 1.98% of slide height; #FFFFFF; center aligned]: A complete and optimized / platform enables more / customers to deploy AI / factories and expands / NVIDIA content per factory. (Line spacing about 2.4% of slide height.)
- body at x 80.2, y 53.2, w 13, h 11.5 [NVIDIASans, sans-serif, Regular weight, normal case; font size 1.98% of slide height; #FFFFFF; center aligned]: Open models, enterprise / agents, sovereign AI and / physical AI unlock new / builders, industries, / and markets. (Line spacing about 2.4% of slide height.)
- shape at x 12.5, y 69.5, w 2, h 2.3 [#76B900]: Card 1 chevron, downward V (Stroke about 3px, no fill, round line caps.)
- shape at x 30.8, y 69.5, w 2, h 2.3 [#76B900]: Card 2 chevron, downward V (Stroke about 3px, no fill, round line caps.)
- shape at x 49.15, y 69.5, w 2, h 2.3 [#76B900]: Card 3 chevron, downward V (Stroke about 3px, no fill, round line caps.)
- shape at x 67.45, y 69.5, w 2, h 2.3 [#76B900]: Card 4 chevron, downward V (Stroke about 3px, no fill, round line caps.)
- shape at x 85.8, y 69.5, w 2, h 2.3 [#76B900]: Card 5 chevron, downward V (Stroke about 3px, no fill, round line caps.)
- body at x 6.4, y 74.4, w 14.2, h 9.1 [NVIDIASans, sans-serif, Regular weight, normal case; font size 1.98% of slide height; #FFFFFF; center aligned]: Existing customer demand / growing >100% year over / year, with an expanding / customer base. (Line spacing about 2.4% of slide height.)
- body at x 25.1, y 74.4, w 13.2, h 4.4 [NVIDIASans, sans-serif, Regular weight, normal case; font size 1.98% of slide height; #FFFFFF; center aligned]: Vera Rubin delivers up to / 45X lower cost per token. (Line spacing about 2.4% of slide height.)
- body at x 43, y 74.4, w 14.2, h 9.1 [NVIDIASans, sans-serif, Regular weight, normal case; font size 1.98% of slide height; #FFFFFF; center aligned]: ~10 years A100 productive / life. Excellent cloud rental / economics and strong / residual value. (Line spacing about 2.4% of slide height.)
- body at x 61.9, y 74.4, w 13, h 6.8 [NVIDIASans, sans-serif, Regular weight, normal case; font size 1.98% of slide height; #FFFFFF; center aligned]: NVIDIA content growing: / Hopper $18B/GW to / Rubin $40B/GW. (Line spacing about 2.4% of slide height.)
- body at x 80.2, y 74.4, w 13.3, h 6.8 [NVIDIASans, sans-serif, Regular weight, normal case; font size 1.98% of slide height; #FFFFFF; center aligned]: ACI&E approx. half of the / business and growing / ~100% year over year. (Line spacing about 2.4% of slide height.)
- logo at x 92.85, y 94.1, w 4.65, h 1.8 [#76B900]: NVIDIA logo: eye symbol with NVIDIA wordmark (Green #76B900 eye mark at left with white bold NVIDIA wordmark at right; sits at bottom right corner, no overlap.)

**Measured from the file** (exact, read from the PDF rather than estimated; x, y, w, h as % of the slide; font size as % of the slide height)

- fill at x 0.0, y 0.0, w 100.0, h 100.0: #000000 (covers the slide)
- text at x 21.5, y 10.5, w 57.0, h 4.9 [NVIDIASans-Medium; font size 4.94; #FFFFFF]: NVIDIA’s Unique Position in the AI Economy
- text at x 29.1, y 18.1, w 42.0, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: NVIDIA combines CUDA programmable compute, full-stack AI factory co-design,
- text at x 26.9, y 20.7, w 46.2, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: a developer ecosystem and global deployment channels in one AI infrastructure platform.
- text at x 6.5, y 46.9, w 13.6, h 2.5 [NVIDIASans-Medium; font size 2.47; #76B900]: AI Workload Breadth
- text at x 25.3, y 46.9, w 12.8, h 2.5 [NVIDIASans-Medium; font size 2.47; #76B900]: Extreme Co-Design
- text at x 44.3, y 46.9, w 11.3, h 2.5 [NVIDIASans-Medium; font size 2.47; #76B900]: CUDA Ecosystem
- text at x 61.5, y 46.9, w 13.7, h 2.5 [NVIDIASans-Medium; font size 2.47; #76B900]: Full-Stack AI Factory
- text at x 79.9, y 46.9, w 13.6, h 2.5 [NVIDIASans-Medium; font size 2.47; #76B900]: Global Market Reach
- text at x 8.4, y 53.2, w 10.3, h 2.0 [NVIDIASans-Regular; font size 1.98; #FFFFFF]: One programmable
- text at x 25.0, y 53.2, w 13.7, h 2.0 [NVIDIASans-Regular; font size 1.98; #FFFFFF]: Compute, networking and
- text at x 42.6, y 53.2, w 15.1, h 2.0 [NVIDIASans-Regular; font size 1.98; #FFFFFF]: CUDA compounds developer
- text at x 61.6, y 53.2, w 13.8, h 2.0 [NVIDIASans-Regular; font size 1.98; #FFFFFF]: A complete and optimized
- text at x 80.3, y 53.2, w 13.0, h 2.0 [NVIDIASans-Regular; font size 1.98; #FFFFFF]: Open models, enterprise
- text at x 7.1, y 55.6, w 12.7, h 2.0 [NVIDIASans-Regular; font size 1.98; #FFFFFF]: platform spans training,
- text at x 26.0, y 55.6, w 11.7, h 2.0 [NVIDIASans-Regular; font size 1.98; #FFFFFF]: software co-designed
- text at x 43.6, y 55.6, w 12.7, h 2.0 [NVIDIASans-Regular; font size 1.98; #FFFFFF]: innovation and installed-
- text at x 62.4, y 55.6, w 12.1, h 2.0 [NVIDIASans-Regular; font size 1.98; #FFFFFF]: platform enables more
- text at x 80.3, y 55.6, w 12.9, h 2.0 [NVIDIASans-Regular; font size 1.98; #FFFFFF]: agents, sovereign AI and
- text at x 7.1, y 58.0, w 12.8, h 2.0 [NVIDIASans-Regular; font size 1.98; #FFFFFF]: inference and agentic AI
- text at x 25.3, y 58.0, w 12.5, h 2.0 [NVIDIASans-Regular; font size 1.98; #FFFFFF]: at factory scale improve
- text at x 43.0, y 58.0, w 14.3, h 2.0 [NVIDIASans-Regular; font size 1.98; #FFFFFF]: base value, keeping NVIDIA
- text at x 62.3, y 58.0, w 12.3, h 2.0 [NVIDIASans-Regular; font size 1.98; #FFFFFF]: customers to deploy AI
- text at x 80.8, y 58.0, w 11.9, h 2.0 [NVIDIASans-Regular; font size 1.98; #FFFFFF]: physical AI unlock new
- text at x 7.3, y 60.3, w 12.4, h 2.0 [NVIDIASans-Regular; font size 1.98; #FFFFFF]: across open and closed
- text at x 26.4, y 60.3, w 10.7, h 2.0 [NVIDIASans-Regular; font size 1.98; #FFFFFF]: tokens per watt and
- text at x 44.7, y 60.3, w 11.0, h 2.0 [NVIDIASans-Regular; font size 1.98; #FFFFFF]: compute productive,
- text at x 62.6, y 60.3, w 11.8, h 2.0 [NVIDIASans-Regular; font size 1.98; #FFFFFF]: factories and expands
- text at x 81.6, y 60.3, w 10.4, h 2.0 [NVIDIASans-Regular; font size 1.98; #FFFFFF]: builders, industries,
- text at x 6.5, y 62.7, w 13.9, h 2.0 [NVIDIASans-Regular; font size 1.98; #FFFFFF]: models, capturing more of
- text at x 27.2, y 62.7, w 8.9, h 2.0 [NVIDIASans-Regular; font size 1.98; #FFFFFF]: tokens per dollar.
- text at x 44.6, y 62.7, w 10.9, h 2.0 [NVIDIASans-Regular; font size 1.98; #FFFFFF]: fungible and durable.
- text at x 61.2, y 62.7, w 14.3, h 2.0 [NVIDIASans-Regular; font size 1.98; #FFFFFF]: NVIDIA content per factory.
- text at x 83.3, y 62.7, w 6.8, h 2.0 [NVIDIASans-Regular; font size 1.98; #FFFFFF]: and markets.
- text at x 6.7, y 65.1, w 13.3, h 2.0 [NVIDIASans-Regular; font size 1.98; #FFFFFF]: AI’s expanding workloads.
- text at x 6.4, y 74.4, w 14.2, h 2.0 [NVIDIASans-Regular; font size 1.98; #FFFFFF]: Existing customer demand
- text at x 25.3, y 74.4, w 13.1, h 2.0 [NVIDIASans-Regular; font size 1.98; #FFFFFF]: Vera Rubin delivers up to
- text at x 43.0, y 74.4, w 14.2, h 2.0 [NVIDIASans-Regular; font size 1.98; #FFFFFF]: ~10 years A100 productive
- text at x 61.9, y 74.4, w 13.0, h 2.0 [NVIDIASans-Regular; font size 1.98; #FFFFFF]: NVIDIA content growing:
- text at x 80.2, y 74.4, w 13.3, h 2.0 [NVIDIASans-Regular; font size 1.98; #FFFFFF]: ACI&E approx. half of the
- text at x 6.8, y 76.8, w 13.3, h 2.0 [NVIDIASans-Regular; font size 1.98; #FFFFFF]: growing >100% year over
- text at x 25.1, y 76.8, w 13.2, h 2.0 [NVIDIASans-Regular; font size 1.98; #FFFFFF]: 45X lower cost per token.
- text at x 43.4, y 76.8, w 13.4, h 2.0 [NVIDIASans-Regular; font size 1.98; #FFFFFF]: life. Excellent cloud rental
- text at x 63.0, y 76.8, w 10.9, h 2.0 [NVIDIASans-Regular; font size 1.98; #FFFFFF]: Hopper $18B/GW to
- text at x 81.0, y 76.8, w 11.5, h 2.0 [NVIDIASans-Regular; font size 1.98; #FFFFFF]: business and growing
- text at x 7.3, y 79.2, w 12.5, h 2.0 [NVIDIASans-Regular; font size 1.98; #FFFFFF]: year, with an expanding
- text at x 44.2, y 79.2, w 11.8, h 2.0 [NVIDIASans-Regular; font size 1.98; #FFFFFF]: economics and strong
- text at x 64.1, y 79.2, w 8.4, h 2.0 [NVIDIASans-Regular; font size 1.98; #FFFFFF]: Rubin $40B/GW.
- text at x 81.0, y 79.2, w 11.5, h 2.0 [NVIDIASans-Regular; font size 1.98; #FFFFFF]: ~100% year over year.
- text at x 9.4, y 81.5, w 8.0, h 2.0 [NVIDIASans-Regular; font size 1.98; #FFFFFF]: customer base.
- text at x 46.4, y 81.5, w 7.2, h 2.0 [NVIDIASans-Regular; font size 1.98; #FFFFFF]: residual value.

### Slide 5

**Text**

Since Our Q2 Earnings Announcement…
1.
New models/agents on NVIDIA infrastructure – Astra (OpenAI), Muse (Meta), Grok Bot (SpaceXAI)
2.
NVIDIA announced Open Agent Safety Platform with OpenShell and Sentry with broad industry support
3.
NVIDIA to acquire Hugging Face for $12.9 billion
4.
Anthropic contracts 2.6 GW to date in NVIDIA AI infrastructure to be shipped through 2028
5.
Anthropic’s total reported contracted value across multiple CSPs/neoclouds now exceeds $180B
6.
NVIDIA’s neocloud partners in Australia deliver up to 2GW by 2027
7.
Nscale and Firmus file for IPO
8.
Palantir to deliver Ontology for Cybersecurity via Cisco’s Security AI factory with NVIDIA as a preferred full-stack foundation
9.
d-Matrix to connect its next-gen XPUs to NVIDIA AI infrastructure via NVLink Fusion
10.
Cisco adds rack-scale AI computing solutions to its portfolio

**Layout**

A single centred title sits near the top of a black slide, with ten numbered items in one left-aligned column below it, each item on its own line with even vertical spacing of about 6 percent of the slide height. The numerals sit in a narrow left column at about 5 percent from the left edge, with the item text starting at about 8 percent; the NVIDIA logo sits bottom right, with no footer, slide number or other furniture.

**Visuals**

One small NVIDIA logo at bottom right: the green NVIDIA eye mark (#76B900) followed by the white NVIDIA wordmark in bold capitals. No charts, photographs, diagrams or other graphics.

**Design**

Background: Solid black #000000 covering the whole slide. Solid black background (#000000) with no gradients, shadows or textures. Title in white (#FFFFFF) NVIDIASans Medium, sans-serif, at about 4.9 percent of slide height, centred. Numerals in NVIDIA green (#76B900) and item text in light grey (#A1A1A1), both NVIDIASans Regular, sans-serif, at about 2.5 percent of slide height, left-aligned. No line or shape styles beyond the logo.

**Elements and positions** (back to front; x, y, width, height as % of the slide; type, colour and alignment in brackets)

- shape at x 0, y 0, w 100, h 100 [#000000]: Full-slide black background rectangle (Covers the entire slide, no border, no shadow.)
- title at x 23.9, y 10.5, w 52.1, h 4.9 [NVIDIASans Medium, sans-serif, medium weight, no italics, no caps, normal letter-spacing; font size 4.94% of slide height; #FFFFFF; center aligned]: Since Our Q2 Earnings Announcement…
- body at x 5.3, y 26, w 1.1, h 2.5 [NVIDIASans Regular, sans-serif, regular weight, no italics, no caps, normal letter-spacing; font size 2.47% of slide height; #76B900; left aligned]: 1. (Numbered list marker, period-style numerals.)
- body at x 7.8, y 26, w 63, h 2.5 [NVIDIASans Regular, sans-serif, regular weight, no italics, no caps, normal letter-spacing; font size 2.47% of slide height; #A1A1A1; left aligned]: New models/agents on NVIDIA infrastructure – Astra (OpenAI), Muse (Meta), Grok Bot (SpaceXAI)
- body at x 5.3, y 32, w 1.1, h 2.5 [NVIDIASans Regular, sans-serif, regular weight, no italics, no caps, normal letter-spacing; font size 2.47% of slide height; #76B900; left aligned]: 2. (Numbered list marker, period-style numerals.)
- body at x 7.8, y 32, w 67.1, h 2.5 [NVIDIASans Regular, sans-serif, regular weight, no italics, no caps, normal letter-spacing; font size 2.47% of slide height; #A1A1A1; left aligned]: NVIDIA announced Open Agent Safety Platform with OpenShell and Sentry with broad industry support
- body at x 5.3, y 37.9, w 1.1, h 2.5 [NVIDIASans Regular, sans-serif, regular weight, no italics, no caps, normal letter-spacing; font size 2.47% of slide height; #76B900; left aligned]: 3. (Numbered list marker, period-style numerals.)
- body at x 7.8, y 37.9, w 31.2, h 2.5 [NVIDIASans Regular, sans-serif, regular weight, no italics, no caps, normal letter-spacing; font size 2.47% of slide height; #A1A1A1; left aligned]: NVIDIA to acquire Hugging Face for $12.9 billion
- body at x 5.3, y 43.8, w 1.1, h 2.5 [NVIDIASans Regular, sans-serif, regular weight, no italics, no caps, normal letter-spacing; font size 2.47% of slide height; #76B900; left aligned]: 4. (Numbered list marker, period-style numerals.)
- body at x 7.8, y 43.8, w 59.2, h 2.5 [NVIDIASans Regular, sans-serif, regular weight, no italics, no caps, normal letter-spacing; font size 2.47% of slide height; #A1A1A1; left aligned]: Anthropic contracts 2.6 GW to date in NVIDIA AI infrastructure to be shipped through 2028
- body at x 5.3, y 49.7, w 1.1, h 2.5 [NVIDIASans Regular, sans-serif, regular weight, no italics, no caps, normal letter-spacing; font size 2.47% of slide height; #76B900; left aligned]: 5. (Numbered list marker, period-style numerals.)
- body at x 7.8, y 49.7, w 62.6, h 2.5 [NVIDIASans Regular, sans-serif, regular weight, no italics, no caps, normal letter-spacing; font size 2.47% of slide height; #A1A1A1; left aligned]: Anthropic’s total reported contracted value across multiple CSPs/neoclouds now exceeds $180B
- body at x 5.3, y 55.7, w 1.1, h 2.5 [NVIDIASans Regular, sans-serif, regular weight, no italics, no caps, normal letter-spacing; font size 2.47% of slide height; #76B900; left aligned]: 6. (Numbered list marker, period-style numerals.)
- body at x 7.8, y 55.7, w 42.9, h 2.5 [NVIDIASans Regular, sans-serif, regular weight, no italics, no caps, normal letter-spacing; font size 2.47% of slide height; #A1A1A1; left aligned]: NVIDIA’s neocloud partners in Australia deliver up to 2GW by 2027
- body at x 5.3, y 61.6, w 1.1, h 2.5 [NVIDIASans Regular, sans-serif, regular weight, no italics, no caps, normal letter-spacing; font size 2.47% of slide height; #76B900; left aligned]: 7. (Numbered list marker, period-style numerals.)
- body at x 7.8, y 61.6, w 19.2, h 2.5 [NVIDIASans Regular, sans-serif, regular weight, no italics, no caps, normal letter-spacing; font size 2.47% of slide height; #A1A1A1; left aligned]: Nscale and Firmus file for IPO
- body at x 5.3, y 67.5, w 1.1, h 2.5 [NVIDIASans Regular, sans-serif, regular weight, no italics, no caps, normal letter-spacing; font size 2.47% of slide height; #76B900; left aligned]: 8. (Numbered list marker, period-style numerals.)
- body at x 7.8, y 67.5, w 80.4, h 2.5 [NVIDIASans Regular, sans-serif, regular weight, no italics, no caps, normal letter-spacing; font size 2.47% of slide height; #A1A1A1; left aligned]: Palantir to deliver Ontology for Cybersecurity via Cisco’s Security AI factory with NVIDIA as a preferred full-stack foundation
- body at x 5.3, y 73.4, w 1.1, h 2.5 [NVIDIASans Regular, sans-serif, regular weight, no italics, no caps, normal letter-spacing; font size 2.47% of slide height; #76B900; left aligned]: 9. (Numbered list marker, period-style numerals.)
- body at x 7.8, y 73.4, w 54.6, h 2.5 [NVIDIASans Regular, sans-serif, regular weight, no italics, no caps, normal letter-spacing; font size 2.47% of slide height; #A1A1A1; left aligned]: d-Matrix to connect its next-gen XPUs to NVIDIA AI infrastructure via NVLink Fusion
- body at x 5.3, y 79.4, w 2, h 2.5 [NVIDIASans Regular, sans-serif, regular weight, no italics, no caps, normal letter-spacing; font size 2.47% of slide height; #76B900; left aligned]: 10. (Numbered list marker, period-style numerals.)
- body at x 7.8, y 79.4, w 39.1, h 2.5 [NVIDIASans Regular, sans-serif, regular weight, no italics, no caps, normal letter-spacing; font size 2.47% of slide height; #A1A1A1; left aligned]: Cisco adds rack-scale AI computing solutions to its portfolio
- logo at x 92.9, y 94, w 4.6, h 1.8 [#76B900]: NVIDIA logo: green eye mark with white NVIDIA wordmark, bottom right corner (Dominant fill is green #76B900 for the eye mark, wordmark in white #FFFFFF; no background behind it, sits directly on the black slide.)

**Measured from the file** (exact, read from the PDF rather than estimated; x, y, w, h as % of the slide; font size as % of the slide height)

- fill at x 0.0, y 0.0, w 100.0, h 100.0: #000000 (covers the slide)
- text at x 23.9, y 10.5, w 52.1, h 4.9 [NVIDIASans-Medium; font size 4.94; #FFFFFF]: Since Our Q2 Earnings Announcement…
- text at x 5.3, y 26.0, w 1.1, h 2.5 [NVIDIASans-Regular; font size 2.47; #76B900]: 1.
- text at x 7.8, y 26.0, w 63.0, h 2.5 [NVIDIASans-Regular; font size 2.47; #A1A1A1]: New models/agents on NVIDIA infrastructure – Astra (OpenAI), Muse (Meta), Grok Bot (SpaceXAI)
- text at x 5.3, y 32.0, w 1.1, h 2.5 [NVIDIASans-Regular; font size 2.47; #76B900]: 2.
- text at x 7.8, y 32.0, w 67.1, h 2.5 [NVIDIASans-Regular; font size 2.47; #A1A1A1]: NVIDIA announced Open Agent Safety Platform with OpenShell and Sentry with broad industry support
- text at x 5.3, y 37.9, w 1.1, h 2.5 [NVIDIASans-Regular; font size 2.47; #76B900]: 3.
- text at x 7.8, y 37.9, w 31.2, h 2.5 [NVIDIASans-Regular; font size 2.47; #A1A1A1]: NVIDIA to acquire Hugging Face for $12.9 billion
- text at x 5.3, y 43.8, w 1.1, h 2.5 [NVIDIASans-Regular; font size 2.47; #76B900]: 4.
- text at x 7.8, y 43.8, w 59.2, h 2.5 [NVIDIASans-Regular; font size 2.47; #A1A1A1]: Anthropic contracts 2.6 GW to date in NVIDIA AI infrastructure to be shipped through 2028
- text at x 5.3, y 49.7, w 1.1, h 2.5 [NVIDIASans-Regular; font size 2.47; #76B900]: 5.
- text at x 7.8, y 49.7, w 62.6, h 2.5 [NVIDIASans-Regular; font size 2.47; #A1A1A1]: Anthropic’s total reported contracted value across multiple CSPs/neoclouds now exceeds $180B
- text at x 5.3, y 55.7, w 1.1, h 2.5 [NVIDIASans-Regular; font size 2.47; #76B900]: 6.
- text at x 7.8, y 55.7, w 42.9, h 2.5 [NVIDIASans-Regular; font size 2.47; #A1A1A1]: NVIDIA’s neocloud partners in Australia deliver up to 2GW by 2027
- text at x 5.3, y 61.6, w 1.1, h 2.5 [NVIDIASans-Regular; font size 2.47; #76B900]: 7.
- text at x 7.8, y 61.6, w 19.2, h 2.5 [NVIDIASans-Regular; font size 2.47; #A1A1A1]: Nscale and Firmus file for IPO
- text at x 5.3, y 67.5, w 1.1, h 2.5 [NVIDIASans-Regular; font size 2.47; #76B900]: 8.
- text at x 7.8, y 67.5, w 80.4, h 2.5 [NVIDIASans-Regular; font size 2.47; #A1A1A1]: Palantir to deliver Ontology for Cybersecurity via Cisco’s Security AI factory with NVIDIA as a preferred full-stack foundation
- text at x 5.3, y 73.4, w 1.1, h 2.5 [NVIDIASans-Regular; font size 2.47; #76B900]: 9.
- text at x 7.8, y 73.4, w 54.6, h 2.5 [NVIDIASans-Regular; font size 2.47; #A1A1A1]: d-Matrix to connect its next-gen XPUs to NVIDIA AI infrastructure via NVLink Fusion
- text at x 5.3, y 79.4, w 2.0, h 2.5 [NVIDIASans-Regular; font size 2.47; #76B900]: 10.
- text at x 7.8, y 79.4, w 39.1, h 2.5 [NVIDIASans-Regular; font size 2.47; #A1A1A1]: Cisco adds rack-scale AI computing solutions to its portfolio

### Slide 6

**Text**

NVIDIA Announces a $150B Share Repurchase Authorization Increase / NVIDIA makes strategic investments across the ecosystem to develop and accelerate growth of the market and CUDA ecosystem. Today, NVIDIA's portfolio consists of / 13 public companies and 229 private companies. Realized ROI on exits-to-date exceeds 3X. We are committed to sharing the company's success with shareholders and / will return excess free cash flow net of strategic uses in the form of share repurchases and a gradually growing dividend. On September 28, we announced a $150 / billion increase to our share repurchase authorization, bringing our total authorization to $235 billion, which we expect to utilize through FY28. / Strategic Investments Across All Layers of the AI Cake / Applications / ABRIDGE / Agility Robotics / NOKIA / Orbital Industries / Revolut / SYNOPSYS / WAYVE / Models / ANTHROPIC / Mistral AI / OpenAI / SPACEX / ssi / THINKING MACHINES / Infrastructure / CoreWeave / Crusoe / firmus / Lambda / NEBIUS / NSCALE / VOLTA / Chips / COHERENT / CORNING / intel. / LUMENTUM / MARVELL / MEDIATEK / SiFive / Energy / LPS / ATOMIC CANYON / LANCIUM / REDWOOD MATERIALS / SB Energy / TerraPower / Note: Strategic investments are not exclusive to those included in the table. / Increasing Stock Repurchases ($ Billions) / $250 / $235 / $200 / $150 / FY27: $53B YTD / Already / $13B above / full-year FY26 / New Authorization, / $150 / $100 / $50 / $40 / $34 / $10 / $0 / FY24 / FY25 / FY26 / FY27 YTD / Total / Remaining Prior / Authorization, / $85 / Authorization

**Layout**

Black full-bleed slide with a centred single-line title at the top (about 10.5% from the top) spanning nearly the full width, followed by a centred four-line grey intro paragraph. Below, two equal rounded-outline panels sit side by side on a two-column grid: left panel (about x 5 to 49, y 31 to 86) holds the logo grid, right panel (about x 51 to 95, y 31 to 86) holds a bar chart. A small NVIDIA logo sits bottom right at about x 93 to 98, y 94 to 96; no slide number is visible.

**Visuals**

Left panel: a logo matrix in five groups, each with a green category label centred above a row of white logos separated by thin grey horizontal rules. Applications (7 logos): Abridge, Agility Robotics, Nokia, Orbital Industries, Revolut, Synopsys, Wayve. Models (6 logos): Anthropic, Mistral AI, OpenAI, SpaceX, SSI, Thinking Machines. Infrastructure (7 logos): CoreWeave, Crusoe, firmus, Lambda, Nebius, Nscale, Volta. Chips (7 logos): Coherent, Corning, Intel, Lumentum, Marvell, MediaTek, SiFive. Energy / LPS (5 logos): Atomic Canyon, Lancium, Redwood Materials, SB Energy, TerraPower. All logos are white or light grey on black. Right panel: a vertical bar chart titled 'Increasing Stock Repurchases ($ Billions)', y-axis $0 to $250 in steps of $50 with labels, x-axis categories FY24, FY25, FY26, FY27 YTD, Total Authorization. Series values as printed: FY24 $10, FY25 $34, FY26 $40, FY27 YTD $53 (highlighted in bright lime #B8FF3C with a callout box), Total Authorization $235 (stacked: New Authorization $150 in deep purple #351147 on top, Remaining Prior Authorization $85 in darker purple at the base). Bars for FY24 to FY26 are dark olive green #3B5D00; the FY27 YTD bar is lime green. A rounded callout box with lime outline (about x 75 to 90, y 51 to 64) reads 'FY27: $53B YTD / Already $13B above full-year FY26', connected by a thin lime leader line from the box to the top of the FY27 YTD bar. No legend; the note about the stacked total is in-chart text labels.

**Design**

Background: Solid black #000000 covering the entire slide. Pure black background (#000000). Titles and primary text white (#FFFFFF); intro paragraph and footnote grey (#A1A1A1 and #757575); accent category labels and highlights NVIDIA green (#76B900) and lime (#B8FF3C). Typeface NVIDIASans, a geometric grotesque sans-serif: Medium for titles and category labels, Regular for body, Bold for highlighted figures. Title is about 4.9% of slide height, intro and chart labels about 1.7 to 2.0%. Panels have thin grey (#5A5A5A approx.) rounded outlines with small corner radius and no fill or shadow. Bars are flat with no gradient.

**Elements and positions** (back to front; x, y, width, height as % of the slide; type, colour and alignment in brackets)

- shape at x 0, y 0, w 100, h 100 [#000000]: Full-slide background rectangle (Covers the slide; no border)
- shape at x 4.9, y 31, w 44.3, h 54.8 [#000000]: Left panel outline (logo grid container) (No fill; thin grey rounded outline about 1 px, corner radius about 1% of slide width)
- shape at x 50.7, y 31, w 44.3, h 54.8 [#000000]: Right panel outline (chart container) (No fill; thin grey rounded outline about 1 px, corner radius about 1% of slide width)
- title at x 4.4, y 10.5, w 91, h 4.9 [NVIDIASans Medium; font size 4.94% of slide height; #FFFFFF; center aligned]: NVIDIA Announces a $150B Share Repurchase Authorization Increase (Single line, sentence case)
- body at x 6.8, y 18.1, w 86.6, h 8.6 [NVIDIASans Regular; font size 1.98% of slide height; #A1A1A1; center aligned]: NVIDIA makes strategic investments across the ecosystem to develop and accelerate growth of the market and CUDA ecosystem. Today, NVIDIA's portfolio consists of / 13 public companies and 229 private companies. Realized ROI on exits-to-date exceeds 3X. We are committed to sharing the company's success with shareholders and / will return excess free cash flow net of strategic uses in the form of share repurchases and a gradually growing dividend. On September 28, we announced a $150 / billion increase to our share repurchase authorization, bringing our total authorization to $235 billion, which we expect to utilize through FY28. (Four centred lines, line pitch about 2.6% of slide height)
- subtitle at x 12.8, y 34.3, w 28.7, h 2 [NVIDIASans Medium; font size 1.98% of slide height; #FFFFFF; center aligned]: Strategic Investments Across All Layers of the AI Cake
- subtitle at x 62.1, y 34.3, w 21.5, h 2 [NVIDIASans Medium; font size 1.98% of slide height; #FFFFFF; center aligned]: Increasing Stock Repurchases ($ Billions)
- body at x 24.6, y 40.5, w 4.9, h 1.5 [NVIDIASans Medium; font size 1.48% of slide height; #76B900; center aligned]: Applications
- logo at x 8.5, y 43, w 37.5, h 4 [#FFFFFF]: Applications row: Abridge, Agility Robotics, Nokia, Orbital Industries, Revolut, Synopsys, Wayve (White logos on black, left to right in that order; thin grey rule below at about y 53.8)
- body at x 25.6, y 49.8, w 2.9, h 1.5 [NVIDIASans Medium; font size 1.48% of slide height; #76B900; center aligned]: Models
- logo at x 10, y 51.8, w 35, h 3.5 [#FFFFFF]: Models row: Anthropic, Mistral AI, OpenAI, SpaceX, SSI, Thinking Machines (White logos on black, left to right; thin grey rule below at about y 62.5)
- body at x 24.2, y 57.5, w 5.6, h 1.5 [NVIDIASans Medium; font size 1.48% of slide height; #76B900; center aligned]: Infrastructure
- logo at x 8.5, y 59.9, w 38, h 4 [#FFFFFF]: Infrastructure row: CoreWeave, Crusoe, firmus, Lambda, Nebius, Nscale, Volta (White logos on black, left to right; thin grey rule below at about y 71.5)
- body at x 25.9, y 65.1, w 2.2, h 1.5 [NVIDIASans Medium; font size 1.48% of slide height; #76B900; center aligned]: Chips
- logo at x 9.5, y 66.8, w 36.5, h 3 [#FFFFFF]: Chips row: Coherent, Corning, Intel, Lumentum, Marvell, MediaTek, SiFive (White logos on black, left to right; thin grey rule below at about y 80.3)
- body at x 24.5, y 73.2, w 5, h 1.5 [NVIDIASans Medium; font size 1.48% of slide height; #76B900; center aligned]: Energy / LPS
- logo at x 9.3, y 75.5, w 36, h 3.5 [#FFFFFF]: Energy / LPS row: Atomic Canyon, Lancium, Redwood Materials, SB Energy, TerraPower (White logos on black, left to right)
- caption at x 12.3, y 82.9, w 29.8, h 1.5 [NVIDIASans Regular; font size 1.48% of slide height; #757575; center aligned]: Note: Strategic investments are not exclusive to those included in the table.
- chart at x 52, y 36, w 42, h 46 [#3B5D00]: Bar chart: Increasing Stock Repurchases ($ Billions) (Y axis $0 to $250 with labels at $0, $50, $100, $150, $200, $250; X categories FY24, FY25, FY26, FY27 YTD, Total Authorization; values $10, $34, $40, $53, $235 printed above bars; FY24 to FY26 bars dark olive #3B5D00; FY27 YTD bar lime; total bar stacked with New Authorization $150 (purple #351147) above Remaining Prior Authorization $85 (darker purple); thin white baseline at $0)
- shape at x 59.3, y 70.5, w 16.4, h 6.1 [#3B5D00]: Filled bar (dark olive green) (Flat fill, no outline)
- shape at x 86.5, y 40.6, w 2.7, h 22.9 [#351147]: Stacked total bar upper segment: New Authorization $150 (Flat fill, no outline)
- shape at x 75, y 50.9, w 12, h 13 [#000000]: Callout box for FY27 YTD (Black fill with lime #B8FF3C outline about 2 px, corner radius about 1.5% of slide width; connected by thin lime leader line to top of FY27 YTD bar)
- logo at x 92.9, y 94, w 6.2, h 1.8 [#76B900]: NVIDIA logo (bottom right) (Green eye emblem with NVIDIA wordmark in white)

**Measured from the file** (exact, read from the PDF rather than estimated; x, y, w, h as % of the slide; font size as % of the slide height)

- fill at x 0.0, y 0.0, w 100.0, h 100.0: #000000 (covers the slide)
- fill at x 59.3, y 70.5, w 16.4, h 6.1: #3B5D00
- fill at x 86.5, y 40.6, w 2.7, h 22.9: #351147
- image at x 9.6, y 75.8, w 7.8, h 2.4: 773 by 132 pixels
- image at x 77.5, y 64.0, w 2.4, h 7.6: 68 by 123 pixels
- image at x 77.5, y 64.0, w 2.4, h 7.6: 68 by 123 pixels
- image at x 13.9, y 43.0, w 4.8, h 3.7: 600 by 257 pixels
- image at x 8.7, y 59.9, w 6.1, h 2.1: 727 by 141 pixels
- image at x 33.9, y 75.6, w 5.2, h 2.4: 500 by 131 pixels
- image at x 19.6, y 75.8, w 6.0, h 1.9: 400 by 73 pixels
- image at x 29.9, y 67.2, w 4.7, h 2.4: 280 by 81 pixels
- image at x 41.2, y 75.1, w 4.0, h 2.5: 529 by 188 pixels
- image at x 25.2, y 43.6, w 4.1, h 2.4: 246 by 81 pixels
- image at x 41.0, y 51.8, w 3.4, h 2.5: 202 by 86 pixels
- image at x 41.9, y 43.7, w 3.8, h 2.2: 227 by 76 pixels
- image at x 23.6, y 67.3, w 4.8, h 1.7: 782 by 160 pixels
- image at x 17.0, y 52.1, w 5.3, h 1.5: 318 by 51 pixels
- image at x 40.8, y 67.4, w 3.6, h 2.1: 217 by 73 pixels
- image at x 31.0, y 43.6, w 3.8, h 2.0: 227 by 67 pixels
- image at x 27.7, y 76.0, w 4.0, h 1.8: 239 by 61 pixels
- image at x 31.6, y 60.3, w 3.5, h 1.7: 417 by 115 pixels
- image at x 36.9, y 51.8, w 2.4, h 2.5: 283 by 168 pixels
- image at x 36.5, y 44.0, w 3.8, h 1.5: 227 by 50 pixels
- image at x 24.3, y 52.3, w 3.4, h 1.6: 203 by 55 pixels
- image at x 36.1, y 67.6, w 3.2, h 1.6: 507 by 145 pixels
- image at x 16.6, y 60.5, w 3.5, h 1.5: 421 by 100 pixels
- text at x 4.4, y 10.5, w 91.0, h 4.9 [NVIDIASans-Medium; font size 4.94; #FFFFFF]: NVIDIA Announces a $150B Share Repurchase Authorization Increase​
- text at x 6.8, y 18.1, w 86.6, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: NVIDIA makes strategic investments across the ecosystem to develop and accelerate growth of the market and CUDA ecosystem. Today, NVIDIA’s portfolio consists of
- text at x 6.7, y 20.7, w 86.8, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: 13 public companies and 229 private companies. Realized ROI on exits-to-date exceeds 3X. We are committed to sharing the company’s success with shareholders and
- text at x 8.2, y 23.3, w 83.9, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: will return excess free cash flow net of strategic uses in the form of share repurchases and a gradually growing dividend. On September 28, we announced a $150
- text at x 13.3, y 25.9, w 73.4, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: billion increase to our share repurchase authorization, bringing our total authorization to $235 billion, which we expect to utilize through FY28.
- text at x 12.8, y 34.3, w 28.7, h 2.0 [NVIDIASans-Medium; font size 1.98; #FFFFFF]: Strategic Investments Across All Layers of the AI Cake
- text at x 62.1, y 34.3, w 21.5, h 2.0 [NVIDIASans-Medium; font size 1.98; #FFFFFF]: Increasing Stock Repurchases ($ Billions)
- text at x 53.9, y 37.5, w 2.3, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: $250
- text at x 86.7, y 38.5, w 2.3, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: $235
- text at x 53.9, y 45.1, w 2.3, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: $200
- text at x 24.6, y 40.5, w 4.9, h 1.5 [NVIDIASans-Medium; font size 1.48; #76B900]: Applications
- text at x 53.9, y 52.8, w 2.3, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: $150
- text at x 73.9, y 53.3, w 7.4, h 1.7 [NVIDIASans-Medium; font size 1.73; #FFFFFF]: FY27: $53B YTD
- text at x 25.6, y 49.8, w 2.9, h 1.5 [NVIDIASans-Medium; font size 1.48; #76B900]: Models
- text at x 74.8, y 58.2, w 5.4, h 1.7 [NVIDIASans-Bold; font size 1.73; #B8FF3C]: $13B above
- text at x 84.2, y 50.5, w 7.6, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: New Authorization,
- text at x 53.9, y 60.4, w 2.3, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: $100
- text at x 86.9, y 52.4, w 2.0, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: $150
- text at x 76.1, y 56.3, w 2.9, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Already
- text at x 80.2, y 66.3, w 1.7, h 1.7 [NVIDIASans-Bold; font size 1.73; #B8FF3C]: $53
- text at x 24.2, y 57.5, w 5.6, h 1.5 [NVIDIASans-Medium; font size 1.48; #76B900]: Infrastructure
- text at x 54.5, y 68.1, w 1.7, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: $50
- text at x 73.4, y 68.3, w 1.7, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: $40
- text at x 66.6, y 69.3, w 1.7, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: $34
- text at x 74.8, y 60.2, w 5.6, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: full-year FY26
- text at x 59.7, y 73.0, w 1.7, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: $10
- text at x 25.9, y 65.1, w 2.2, h 1.5 [NVIDIASans-Medium; font size 1.48; #76B900]: Chips
- text at x 55.1, y 75.7, w 1.2, h 1.9 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: $0-
- text at x 59.5, y 78.0, w 2.3, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: FY24
- text at x 66.3, y 78.0, w 2.3, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: FY25
- text at x 73.1, y 78.0, w 2.3, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: FY26
- text at x 78.9, y 78.0, w 4.4, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: FY27 YTD
- text at x 86.7, y 78.0, w 2.3, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: Total
- text at x 84.8, y 67.5, w 6.4, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Remaining Prior
- text at x 84.8, y 80.2, w 6.2, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: Authorization
- text at x 85.2, y 69.4, w 5.7, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Authorization,
- text at x 87.1, y 71.3, w 1.5, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: $85
- text at x 24.5, y 73.2, w 5.0, h 1.5 [NVIDIASans-Medium; font size 1.48; #76B900]: Energy / LPS
- text at x 12.3, y 82.9, w 29.8, h 1.5 [NVIDIASans-Regular; font size 1.48; #757575]: Note: Strategic investments are not exclusive to those included in the table.

### Slide 7

**Text**

NVIDIA Full-Stack AI Factory Platform,
Not the Chip, Determines Economic Output
What NVIDIA builds is fundamentally different from
what other chip companies build.
The Vera Rubin platform integrates seven purpose-
built chips across five rack-scale systems. And,
through the DSX AI factory platform, NVIDIA helps
design, build and operate efficient, resilient and
grid-ready AI factories.
The ability to extreme co-design at AI factory scale
is unique to NVIDIA, and the result is integrated yet
customizable infrastructure that can be deployed
across every major cloud, inside enterprise data
centers, at the edge, and into robotics and physical
AI systems.
NVIDIA delivers x-factor performance gains every year
and continuously improves the performance of its
installed base through software enhancements.
The NVIDIA Full-Stack AI Factory
CUDA-X Libraries
Platform Libraries
Open Models
cuDNN
NIXL
CCCL
DOCA
cuSPARSE
NVSHMEM
CUDA Python
NCCL
NVTX
cuBLAS
TensorRT-LLM
cuTILE Python
Megatron Core
NVRTC
Transformer Engine
CUTLASS
Dynamo
CuTe DSL
Vera Rubin NVL72 Compute Tray
Vera Rubin NVL72 NVLink Switch Tray
Vera CPU Tray
Spectrum-6 SPX Switch Tray
Groq 3 LPX Tray
Vera BlueField-4 STX Storage Tray
Vera CPU Tray
Rubin GPU
Vera
Grace BlueField
ConnectX-9
NVLink 6 Switch
Vera
Spectrum-6 Switch
Groq 3
Vera BlueField-4 STX
Vera BlueField-4
NVIDIA DSX

**Layout**

Single-slide black layout with a centered two-line title across the top, a left text column of body copy starting at about 31% of slide height, and a bordered framed graphic panel occupying the right half from about 38% to 95% horizontally and 28.5% to 85.6% vertically. The left column and panel are roughly top-aligned at the content area, with the left text column left-aligned at a margin of about 5.2% of slide width. The NVIDIA logo sits bottom right at about 93% x and 94.5% y. No footer text and no slide number are visible.

**Visuals**

A single framed panel image (about 57.3% by 57.3% of slide, 16:9 image of 3840 by 2160 px) with a near-black background and a thin light grey rounded border. Its title reads The NVIDIA Full-Stack AI Factory in white sans-serif, centered near the top. The upper part has three groups of icons with captions. Left group, captioned CUDA-X Libraries in gold, shows a row of six overlapping rectangular tiles with green, teal and blue computer-graphics renderings, the central tile brightest with a glowing blue and green circuit image. Centre group, captioned Platform Libraries in gold, is a grid of labelled grey rounded rectangles: top row cuDNN; second row NIXL, CCCL, DOCA; third row cuSPARSE, NVSHMEM, CUDA Python, NCCL, NVTX; fourth row cuBLAS, TensorRT-LLM, cuTILE Python, Megatron Core, NVRTC; fifth row Transformer Engine, CUTLASS, Dynamo; sixth row CuTe DSL. Right group, captioned Open Models in gold, shows six overlapping tiles with figurative renderings: a globe, a DNA-like helix, a central humanoid figure, a spiral and a human body figure. Below is a row of seven line-drawn server rack illustrations in glowing gold and white outline, each with a small two-line caption above: Vera Rubin NVL72 Compute Tray (with labels Rubin GPU, Vera, Grace BlueField, ConnectX-9 pointing to the rack), Vera Rubin NVL72 NVLink Switch Tray (label NVLink 6 Switch), Vera CPU Tray (label Vera), Spectrum-6 SPX Switch Tray (label Spectrum-6 Switch), Groq 3 LPX Tray (label Groq 3), Vera BlueField-4 STX Storage Tray (labels Vera BlueField-4 STX), and Vera CPU Tray (label Vera BlueField-4). The lower band of the panel, on a dark purple-to-blue gradient, shows a large isometric rendering of a flat AI data-centre building with glowing warm white lights along its edges and a blue-lit central hall, captioned NVIDIA DSX in gold at the bottom centre. The panel has no overlay beyond the dark tint and is boxed, not cropped or full-bleed.

**Design**

Background: Solid black #000000 with no gradient, photograph or pattern. Solid black background. Titles are white #FFFFFF in NVIDIASans Medium (a geometric humanist sans-serif), regular case, about 4.9% of slide height per line, with the two lines of the title centred. Body text is light grey #A1A1A1 in NVIDIASans Regular, sans-serif, about 2.0% of slide height, with line spacing of about 3% of slide height and paragraph gaps of about 4%. Accent colour is NVIDIA green #76B900 for the logo mark; gold labels inside the panel are about #D4B25A. The image panel has a thin grey #3A3A3A rounded border with a corner radius of about 1% of slide width. No shadows or gradients outside the panel image.

**Elements and positions** (back to front; x, y, width, height as % of the slide; type, colour and alignment in brackets)

- shape at x 0, y 0, w 100, h 100 [#000000]: Full-slide solid black background fill (Covers entire slide, no border)
- photo at x 37.8, y 28.5, w 57.3, h 57.3 [#000000]: Framed illustration panel: The NVIDIA Full-Stack AI Factory diagram with CUDA-X Libraries, Platform Libraries, Open Models, seven server tray renderings and the NVIDIA DSX data-centre rendering (Boxed 16:9 image, not cropped, not full-bleed; thin grey #3A3A3A rounded border, corner radius about 1% of slide width; dark tint over the image; no shadow)
- title at x 21.7, y 5.2, w 57, h 10.2 [NVIDIASans Medium, regular case; font size 4.94% of slide height; #FFFFFF; center aligned]: NVIDIA Full-Stack AI Factory Platform, / Not the Chip, Determines Economic Output (Two lines, line spacing about 5.3% of slide height)
- body at x 5.2, y 30.9, w 28.4, h 52.4 [NVIDIASans Regular, regular weight; font size 1.98% of slide height; #A1A1A1; left aligned]: What NVIDIA builds is fundamentally different from / what other chip companies build. / The Vera Rubin platform integrates seven purpose- / built chips across five rack-scale systems. And, / through the DSX AI factory platform, NVIDIA helps / design, build and operate efficient, resilient and / grid-ready AI factories. / The ability to extreme co-design at AI factory scale / is unique to NVIDIA, and the result is integrated yet / customizable infrastructure that can be deployed / across every major cloud, inside enterprise data / centers, at the edge, and into robotics and physical / AI systems. / NVIDIA delivers x-factor performance gains every year / and continuously improves the performance of its / installed base through software enhancements. (Paragraphs separated by extra space; line spacing about 3% of slide height; gap before 'NVIDIA delivers' paragraph larger)
- logo at x 92.7, y 94.4, w 6.8, h 2 [#76B900]: NVIDIA logo: green eye mark with NVIDIA wordmark (Green mark #76B900 with white #FFFFFF wordmark, bottom right corner)

**Measured from the file** (exact, read from the PDF rather than estimated; x, y, w, h as % of the slide; font size as % of the slide height)

- fill at x 0.0, y 0.0, w 100.0, h 100.0: #000000 (covers the slide)
- image at x 37.8, y 28.5, w 57.3, h 57.3: 3840 by 2160 pixels
- text at x 24.7, y 5.2, w 51.2, h 4.9 [NVIDIASans-Medium; font size 4.94; #FFFFFF]: NVIDIA Full-Stack AI Factory Platform,
- text at x 21.7, y 10.5, w 57.0, h 4.9 [NVIDIASans-Medium; font size 4.94; #FFFFFF]: Not the Chip, Determines Economic Output
- text at x 5.2, y 30.9, w 27.0, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: What NVIDIA builds is fundamentally different from
- text at x 5.2, y 33.9, w 17.4, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: what other chip companies build.
- text at x 5.2, y 38.3, w 26.4, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: The Vera Rubin platform integrates seven purpose-
- text at x 5.2, y 41.3, w 24.6, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: built chips across five rack-scale systems. And,
- text at x 5.2, y 44.2, w 26.5, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: through the DSX AI factory platform, NVIDIA helps
- text at x 5.2, y 47.2, w 24.8, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: design, build and operate efficient, resilient and
- text at x 5.2, y 50.2, w 12.2, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: grid-ready AI factories.
- text at x 5.2, y 54.6, w 26.7, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: The ability to extreme co-design at AI factory scale
- text at x 5.2, y 57.6, w 26.8, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: is unique to NVIDIA, and the result is integrated yet
- text at x 5.2, y 60.5, w 25.8, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: customizable infrastructure that can be deployed
- text at x 5.2, y 63.5, w 24.9, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: across every major cloud, inside enterprise data
- text at x 5.2, y 66.4, w 26.7, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: centers, at the edge, and into robotics and physical
- text at x 5.2, y 69.4, w 5.9, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: AI systems.
- text at x 5.2, y 75.3, w 28.4, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: NVIDIA delivers x-factor performance gains every year
- text at x 5.2, y 78.3, w 26.2, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: and continuously improves the performance of its
- text at x 5.2, y 81.3, w 24.7, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: installed base through software enhancements.

### Slide 8

**Review notes**

- Review: Weaknesses are consequential: many slides depend on small gray body text and tiny labels (slides 4, 8, 13, 18, 21, 22), the financial tables on slide 9 are unreadable at presentation size, and slide 2 is a dense disclaimer wall.
- Suggested repair: Enlarge or simplify the small body text and diagram labels on slides 4, 8, 13, 18 and 21, and replace the screenshot-style tables on slide 9 with a few legible headline figures.

**Text**

CUDA Programmability and Full-Stack AI Factory
Expand the Global Diverse AI Market
Programmability open new growth | CUDA-X libraries connect ecosystems | Full-stack AI factory open go-to-market options
ENTERPRISE & INDUSTRIAL
CSPs
AI LABS
SELF-HOST
AI NATIVES
NVIDIA
Full-Stack
AI Factory
ODMs & OEMs
NEOCLOUDS
Auto | Gaming | Public Sector
Defense | Healthcare | Retail & Logistics
Energy | M & E | Supercomputing
FSI | Manufacturing | Telco
AWS | Microsoft Azure
Google Cloud | Oracle Cloud
Meta | SpaceXAI
Anthropic
Gemini
Meta
OpenAI
SpaceXAI
10K+ Startups
AI2 | Periodic Labs
AMI Labs | Physical Intelligence
BFL | Prometheus
Cognition | Reflection
Cursor | Revolution Labs
Epochal | Ricursive Intelligence
Flapping Airplanes | River AI
Humans& | Sakana
Ineffable | Skild.AI
Inception Labs | SSI
Magic | TML
Mistral | Cohere
Aivres | Gigabyte | Quanta
Asus | HPE | Supermicro
Bull | Inventec | Wistron
Dell | Lenovo | Wiwynn
Foxconn | Pegatron
CoreWeave | IREN | Reliance
Crusoe | Lambda | Volta
Firebird | Mistral | Yotta
Firmus | Nebius
GMI | Nscale

**Layout**

A centred title block sits at the top with a centred grey subtitle line beneath it. Below, a large donut ring of six green wedges is centred on the slide with a black centre disc, and six dark rounded callout boxes are arranged around it: four on the left and top-left and bottom-left, and three on the right and top-right and bottom-right, each joined to its wedge by a thin line. No footer, page number or logo is visible; margins are roughly 5 percent on each side.

**Visuals**

Donut diagram made of six equal-looking wedges arranged clockwise from the top right: CSPs (top right, pale lime gradient), AI NATIVES (right, light lime gradient), NEOCLOUDS (bottom right, bright lime green gradient), ODMs & OEMs (bottom left, medium green gradient), AI LABS SELF-HOST (left, medium green gradient), ENTERPRISE & INDUSTRIAL (top left, dark olive green gradient). Each wedge is separated from its neighbours by a thin black gap. A black centre disc holds the label NVIDIA Full-Stack AI Factory in white. Six dark callout boxes list the members of each group: Enterprise & Industrial box lists Auto, Defense, Energy, FSI, Gaming, Healthcare, M & E, Manufacturing, Public Sector, Retail & Logistics, Supercomputing, Telco. CSPs box lists AWS, Google Cloud, Meta, Microsoft Azure, Oracle Cloud, SpaceXAI. AI Labs Self-Host box lists Anthropic, Gemini, Meta, OpenAI, SpaceXAI. 10K+ Startups box lists AI2, AMI Labs, BFL, Cognition, Cursor, Epochal, Flapping Airplanes, Humans&, Ineffable, Inception Labs, Magic, Mistral, Periodic Labs, Physical Intelligence, Prometheus, Reflection, Revolution Labs, Ricursive Intelligence, River AI, Sakana, Skild.AI, SSI, TML, Cohere. ODMs & OEMs box lists Aivres, Asus, Bull, Dell, Foxconn, Gigabyte, HPE, Inventec, Lenovo, Pegatron, Quanta, Supermicro, Wistron, Wiwynn. Neoclouds box lists CoreWeave, Crusoe, Firebird, Firmus, GMI, IREN, Lambda, Mistral, Nebius, Nscale, Reliance, Volta, Yotta. No photographs or icons; no logos.

**Design**

Background: Solid black #000000 covering the whole slide. Dark, high-contrast presentation style on pure black. Title and centre label are white NVIDIASans Medium, sans-serif, with the title about 4.9 percent of slide height and body labels about 1.5 percent. Group headings are bold white or black sans-serif (NVIDIASans Bold, about 1.7 percent). Accent colour is NVIDIA green #76B900 used for the 10K+ Startups heading and border and connector lines. Callout boxes are flat #0D0D0D with a thin grey #3A3A3A hairline border and small corner radius of about 0.5 percent of slide width; the 10K+ Startups box has a green #76B900 border. Wedges have subtle gradients and a light edge highlight; no shadows.

**Elements and positions** (back to front; x, y, width, height as % of the slide; type, colour and alignment in brackets)

- shape at x 0, y 0, w 100, h 100 [#000000]: Full-slide background rectangle, black (Covers entire slide, no border)
- title at x 18.1, y 5.2, w 64.5, h 10.3 [NVIDIASans Medium, sans-serif, normal weight; font size 4.94% of slide height; #FFFFFF; center aligned]: CUDA Programmability and Full-Stack AI Factory / Expand the Global Diverse AI Market (Two lines, tight line spacing)
- subtitle at x 17.4, y 18.1, w 65.2, h 2 [NVIDIASans Regular, sans-serif, normal weight; font size 1.98% of slide height; #A1A1A1; center aligned]: Programmability open new growth | CUDA-X libraries connect ecosystems | Full-stack AI factory open go-to-market options
- illustration at x 49.8, y 29.3, w 16.4, h 23.8 [#E8FFC0]: Donut wedge CSPs, pale lime gradient (Image 896 by 731 px, wedge edge against black)
- illustration at x 54.7, y 43.7, w 11.6, h 28.6 [#C8F57A]: Donut wedge AI NATIVES, light lime gradient (Image 630 by 877 px)
- illustration at x 49.8, y 63.1, w 14.2, h 23.9 [#9BE02A]: Donut wedge NEOCLOUDS, bright lime green gradient (Image 773 by 731 px)
- illustration at x 35.8, y 62.8, w 14.2, h 24.2 [#4A7A0E]: Donut wedge ODMs & OEMs, medium green gradient (Image 773 by 740 px)
- illustration at x 33.8, y 44, w 11.5, h 28.3 [#4E7A10]: Donut wedge AI LABS SELF-HOST, medium green gradient (Image 626 by 869 px)
- illustration at x 36, y 29.2, w 13.8, h 23.9 [#2F4A10]: Donut wedge ENTERPRISE & INDUSTRIAL, dark olive green gradient (Image 753 by 730 px)
- shape at x 44.5, y 48.4, w 11, h 19.3 [#000000]: Centre disc, black, holding the NVIDIA label (Circle, thin black gap separates it from wedges)
- text at x 46.7, y 54, w 7, h 8.4 [NVIDIASans Medium, sans-serif; font size 2.47% of slide height; #FFFFFF; center aligned]: NVIDIA / Full-Stack / AI Factory
- text at x 41.1, y 40.2, w 7.1, h 3.8 [NVIDIASans Bold, sans-serif, capitals; font size 1.73% of slide height; #FFFFFF; center aligned]: ENTERPRISE & INDUSTRIAL
- text at x 54.4, y 41, w 2.4, h 1.7 [NVIDIASans Bold, sans-serif; font size 1.73% of slide height; #000000; center aligned]: CSPs
- text at x 37.5, y 55.9, w 4, h 1.7 [NVIDIASans Bold, sans-serif, capitals; font size 1.73% of slide height; #FFFFFF; center aligned]: AI LABS
- text at x 36.7, y 58, w 5.4, h 1.7 [NVIDIASans Bold, sans-serif, capitals; font size 1.73% of slide height; #FFFFFF; center aligned]: SELF-HOST
- text at x 59, y 55.9, w 4.1, h 3.8 [NVIDIASans Bold, sans-serif, capitals; font size 1.73% of slide height; #000000; center aligned]: AI / NATIVES
- text at x 42.7, y 73.4, w 4, h 3.8 [NVIDIASans Bold, sans-serif; font size 1.73% of slide height; #FFFFFF; center aligned]: ODMs & / OEMs
- text at x 52.8, y 74.2, w 6, h 1.7 [NVIDIASans Bold, sans-serif, capitals; font size 1.73% of slide height; #000000; center aligned]: NEOCLOUDS
- line at x 34.9, y 33.8, w 3.4, h 4.4 [#76B900]: Connector from ENTERPRISE & INDUSTRIAL box to its wedge (Thin 1 px line, green)
- line at x 61.7, y 34, w 3.2, h 4.2 [#E8FFC0]: Connector from CSPs box to its wedge (Thin 1 px line, pale lime)
- line at x 66.2, y 56.4, w 3.3, h 0.3 [#76B900]: Connector from AI NATIVES to 10K+ Startups box (Horizontal line, green, about 2 px thick)
- line at x 30.5, y 56.4, w 3.3, h 0.3 [#76B900]: Connector from AI LABS box to its wedge (Horizontal thin line, green)
- line at x 34.9, y 78, w 3.4, h 4.3 [#76B900]: Connector from ODMs & OEMs box to its wedge (Thin 1 px line, green)
- line at x 61.7, y 78, w 3.2, h 4 [#76B900]: Connector from NEOCLOUDS box to its wedge (Thin 1 px line, green)
- shape at x 10.8, y 27.2, w 24, h 13.2 [#0D0D0D]: Callout box, Enterprise & Industrial (Rounded rectangle, grey hairline border #3A3A3A, corner radius about 0.5 percent of width)
- text at x 12.8, y 29.6, w 20.7, h 8.2 [NVIDIASans Regular, sans-serif; font size 1.48% of slide height; #FFFFFF; left aligned]: Auto | Gaming | Public Sector / Defense | Healthcare | Retail & Logistics / Energy | M & E | Supercomputing / FSI | Manufacturing | Telco (Four rows, three columns, line spacing about 2.3 percent of height)
- shape at x 65, y 25.8, w 18.8, h 11.9 [#0D0D0D]: Callout box, CSPs (Rounded rectangle, grey hairline border, corner radius about 0.5 percent of width)
- text at x 67, y 28.6, w 13.9, h 6.2 [NVIDIASans Regular, sans-serif; font size 1.48% of slide height; #FFFFFF; left aligned]: AWS | Microsoft Azure / Google Cloud | Oracle Cloud / Meta | SpaceXAI (Three rows, two columns)
- shape at x 21.6, y 49.3, w 9, h 14.7 [#0D0D0D]: Callout box, AI Labs Self-Host (Rounded rectangle, grey hairline border, corner radius about 0.5 percent of width)
- text at x 23.7, y 51, w 3.8, h 10.9 [NVIDIASans Regular, sans-serif; font size 1.48% of slide height; #FFFFFF; left aligned]: Anthropic / Gemini / Meta / OpenAI / SpaceXAI (Five rows, one column)
- shape at x 69.5, y 40.5, w 19.4, h 35.8 [#0D0D0D]: Callout box, 10K+ Startups (Rounded rectangle, green #76B900 border about 1 px, corner radius about 0.5 percent of width)
- text at x 74.7, y 42.6, w 8.6, h 2.2 [NVIDIASans Medium, sans-serif; font size 2.22% of slide height; #76B900; center aligned]: 10K+ Startups
- text at x 70.7, y 46.7, w 16.9, h 27.3 [NVIDIASans Regular, sans-serif; font size 1.48% of slide height; #FFFFFF; left aligned]: AI2 | Periodic Labs / AMI Labs | Physical Intelligence / BFL | Prometheus / Cognition | Reflection / Cursor | Revolution Labs / Epochal | Ricursive Intelligence / Flapping Airplanes | River AI / Humans& | Sakana / Ineffable | Skild.AI / Inception Labs | SSI / Magic | TML / Mistral | Cohere (Thirteen rows, two columns)
- shape at x 65, y 79.8, w 17.7, h 14.7 [#0D0D0D]: Callout box, Neoclouds (Rounded rectangle, grey hairline border, corner radius about 0.5 percent of width)
- text at x 66.8, y 81.8, w 14.1, h 10.9 [NVIDIASans Regular, sans-serif; font size 1.48% of slide height; #FFFFFF; left aligned]: CoreWeave | IREN | Reliance / Crusoe | Lambda | Volta / Firebird | Mistral | Yotta / Firmus | Nebius / GMI | Nscale (Five rows, three columns)
- shape at x 16, y 76.6, w 18.8, h 16.5 [#0D0D0D]: Callout box, ODMs & OEMs (Rounded rectangle, grey hairline border, corner radius about 0.5 percent of width)
- text at x 17.6, y 79.4, w 15.9, h 10.9 [NVIDIASans Regular, sans-serif; font size 1.48% of slide height; #FFFFFF; left aligned]: Aivres | Gigabyte | Quanta / Asus | HPE | Supermicro / Bull | Inventec | Wistron / Dell | Lenovo | Wiwynn / Foxconn | Pegatron (Five rows, three columns)

**Measured from the file** (exact, read from the PDF rather than estimated; x, y, w, h as % of the slide; font size as % of the slide height)

- fill at x 0.0, y 0.0, w 100.0, h 100.0: #000000 (covers the slide)
- fill at x 69.5, y 40.5, w 19.4, h 35.8: #0D0D0D
- fill at x 65.0, y 79.8, w 17.7, h 14.7: #0D0D0D
- fill at x 65.0, y 25.8, w 18.8, h 11.9: #0D0D0D
- fill at x 21.6, y 49.3, w 9.0, h 14.7: #0D0D0D
- fill at x 16.0, y 76.6, w 18.8, h 16.5: #0D0D0D
- fill at x 10.8, y 27.2, w 24.0, h 13.2: #0D0D0D
- image at x 49.8, y 29.3, w 16.4, h 23.8: 896 by 731 pixels
- image at x 35.8, y 62.8, w 14.2, h 24.2: 773 by 740 pixels
- image at x 49.8, y 63.1, w 14.2, h 23.9: 773 by 731 pixels
- image at x 54.7, y 43.7, w 11.6, h 28.6: 630 by 877 pixels
- image at x 36.0, y 29.2, w 13.8, h 23.9: 753 by 730 pixels
- image at x 33.8, y 44.0, w 11.5, h 28.3: 626 by 869 pixels
- text at x 18.1, y 5.2, w 64.5, h 10.3 [NVIDIASans-Medium; font size 4.94; #FFFFFF]: CUDA Programmability and Full-Stack AI Factory Expand the Global Diverse AI Market
- text at x 17.4, y 18.1, w 65.2, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: Programmability open new growth | CUDA-X libraries connect ecosystems | Full-stack AI factory open go-to-market options
- text at x 74.7, y 42.6, w 8.6, h 2.2 [NVIDIASans-Medium; font size 2.22; #76B900]: 10K+ Startups
- text at x 67.0, y 28.6, w 1.8, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: AWS
- text at x 74.7, y 28.6, w 6.2, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Microsoft Azure
- text at x 12.8, y 29.6, w 1.9, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Auto
- text at x 18.6, y 29.6, w 3.0, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Gaming
- text at x 26.4, y 29.6, w 5.1, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Public Sector
- text at x 67.0, y 31.0, w 5.2, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Google Cloud
- text at x 74.7, y 31.0, w 4.9, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Oracle Cloud
- text at x 12.8, y 31.9, w 3.2, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Defense
- text at x 18.6, y 31.9, w 4.2, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Healthcare
- text at x 26.4, y 31.9, w 6.7, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Retail & Logistics
- text at x 67.0, y 33.3, w 2.0, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Meta
- text at x 74.7, y 33.3, w 3.7, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: SpaceXAI
- text at x 12.8, y 34.3, w 2.7, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Energy
- text at x 18.6, y 34.3, w 2.3, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: M & E
- text at x 26.4, y 34.3, w 6.5, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Supercomputing
- text at x 46.7, y 54.0, w 7.0, h 8.4 [NVIDIASans-Medium; font size 2.47; #FFFFFF]: NVIDIA Full-Stack AI Factory
- text at x 54.4, y 41.0, w 2.4, h 1.7 [NVIDIASans-Bold; font size 1.73; #000000]: CSPs
- text at x 41.1, y 40.2, w 7.1, h 3.8 [NVIDIASans-Bold; font size 1.73; #FFFFFF]: ENTERPRISE & INDUSTRIAL
- text at x 12.8, y 36.6, w 1.2, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: FSI
- text at x 18.6, y 36.6, w 5.8, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Manufacturing
- text at x 26.4, y 36.6, w 2.0, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Telco
- text at x 70.7, y 46.7, w 1.5, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: AI2
- text at x 79.2, y 46.7, w 5.4, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Periodic Labs
- text at x 37.5, y 55.9, w 4.0, h 1.7 [NVIDIASans-Bold; font size 1.73; #FFFFFF]: AI LABS
- text at x 59.0, y 55.9, w 4.1, h 3.8 [NVIDIASans-Bold; font size 1.73; #000000]: AI NATIVES
- text at x 70.7, y 49.1, w 3.8, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: AMI Labs
- text at x 79.2, y 49.1, w 8.3, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Physical Intelligence
- text at x 36.7, y 58.0, w 5.4, h 1.7 [NVIDIASans-Bold; font size 1.73; #FFFFFF]: SELF-HOST
- text at x 23.7, y 51.0, w 3.8, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Anthropic
- text at x 70.7, y 51.4, w 1.7, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: BFL
- text at x 79.2, y 51.4, w 4.9, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Prometheus
- text at x 23.7, y 53.4, w 2.7, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Gemini
- text at x 70.7, y 53.7, w 4.0, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Cognition
- text at x 79.2, y 53.7, w 4.2, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Reflection
- text at x 23.7, y 55.7, w 2.0, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Meta
- text at x 70.7, y 56.1, w 2.8, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Cursor
- text at x 79.2, y 56.1, w 6.4, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Revolution Labs
- text at x 23.7, y 58.0, w 2.9, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: OpenAI
- text at x 70.7, y 58.4, w 3.3, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Epochal
- text at x 79.2, y 58.4, w 8.4, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Ricursive Intelligence
- text at x 23.7, y 60.4, w 3.7, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: SpaceXAI
- text at x 70.7, y 60.8, w 7.2, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Flapping Airplanes
- text at x 79.2, y 60.8, w 3.1, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: River AI
- text at x 70.7, y 63.1, w 4.0, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Humans&
- text at x 79.2, y 63.1, w 3.1, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Sakana
- text at x 42.7, y 73.4, w 4.0, h 3.8 [NVIDIASans-Bold; font size 1.73; #FFFFFF]: ODMs & OEMs
- text at x 52.8, y 74.2, w 6.0, h 1.7 [NVIDIASans-Bold; font size 1.73; #000000]: NEOCLOUDS
- text at x 70.7, y 65.5, w 3.7, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Ineffable
- text at x 79.2, y 65.5, w 3.1, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Skild.AI
- text at x 70.7, y 67.8, w 5.9, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Inception Labs
- text at x 79.2, y 67.8, w 1.5, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: SSI
- text at x 70.7, y 70.1, w 2.4, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Magic
- text at x 79.2, y 70.1, w 1.7, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: TML
- text at x 70.7, y 72.5, w 2.7, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Mistral
- text at x 79.2, y 72.5, w 2.8, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Cohere
- text at x 17.6, y 79.4, w 2.4, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Aivres
- text at x 23.1, y 79.4, w 3.5, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Gigabyte
- text at x 29.0, y 79.5, w 2.9, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Quanta
- text at x 17.6, y 81.8, w 1.9, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Asus
- text at x 23.1, y 81.8, w 1.6, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: HPE
- text at x 29.0, y 81.8, w 4.5, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Supermicro
- text at x 66.8, y 81.8, w 4.4, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: CoreWeave
- text at x 72.6, y 81.8, w 1.9, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: IREN
- text at x 77.7, y 81.8, w 3.2, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Reliance
- text at x 17.6, y 84.1, w 1.4, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Bull
- text at x 23.1, y 84.1, w 3.3, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Inventec
- text at x 29.0, y 84.2, w 3.1, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Wistron
- text at x 66.8, y 84.2, w 2.7, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Crusoe
- text at x 72.6, y 84.2, w 3.1, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Lambda
- text at x 77.7, y 84.2, w 2.0, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Volta
- text at x 17.6, y 86.5, w 1.5, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Dell
- text at x 23.1, y 86.5, w 2.8, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Lenovo
- text at x 29.0, y 86.5, w 3.1, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Wiwynn
- text at x 66.8, y 86.5, w 2.9, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Firebird
- text at x 72.6, y 86.5, w 2.7, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Mistral
- text at x 77.7, y 86.5, w 2.1, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Yotta
- text at x 17.6, y 88.8, w 3.2, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Foxconn
- text at x 23.1, y 88.8, w 3.5, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Pegatron
- text at x 66.8, y 88.9, w 2.6, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Firmus
- text at x 72.6, y 88.9, w 2.7, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Nebius
- text at x 66.8, y 91.2, w 1.6, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: GMI
- text at x 72.6, y 91.2, w 2.6, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Nscale

### Slide 9

**Review notes**

- Review: Weaknesses are consequential: many slides depend on small gray body text and tiny labels (slides 4, 8, 13, 18, 21, 22), the financial tables on slide 9 are unreadable at presentation size, and slide 2 is a dense disclaimer wall.
- Suggested repair: Enlarge or simplify the small body text and diagram labels on slides 4, 8, 13, 18 and 21, and replace the screenshot-style tables on slide 9 with a few legible headline figures.

**Text**

NVIDIA Will Leverage Its Financial Capacity to Support the Ecosystem
Future commitments by fiscal year as of July 26, 2026, were as follows:
Remainder of 2027 / 2028 / 2029 / 2030 / 2031 / 2032 and thereafter / Total
(In billions)
Supply and capacity $ 92 $ 87 $ 88 $ 6 $ 5 $ 1 $ 279
Cloud service agreements 3 8 7 6 4 1 29
Data center leases not commenced — 1 1 2 1 20 25
Equity investments 18 3 2 2 — — 25
Capital expenditures 7 1 — — — — 8
Total $ 120 $ 100 $ 98 $ 16 $ 10 $ 22 $ 366
Future commitments by fiscal year as of July 26, 2026, were as follows:
Remainder of 2027 / 2028 / 2029 / 2030 / 2031 / 2032 and thereafter / Total
(In billions)
AI cloud agreements $ — $ 6 $ 8 $ 7 $ 6 $ 9 $ 36
Data center leases not commenced for third party — — 1 1 1 17 20
Total $ — $ 6 $ 9 $ 8 $ 7 $ 26 $ 56
The following table summarizes the maximum gross exposure related to our guarantees, including the SB Energy Corp. guarantees signed in August 2026 (in billions):
Land, power, and shell guarantees for AI clouds $ 3.5
SB Energy Corp. guarantees 105.0
Total $ 108.5
We’ve partnered with our extensive network of suppliers to secure the critical components needed to meet demand for the next several years. Our commitments increased from $119 billion last quarter to $279 billion, primarily related to the procurement of memory.
We’ve partnered with leading AI clouds to enable broader access to our AI infrastructure to serve AI startups, model builders, enterprises, research organizations and sovereign customers. Under these agreements, we will earn revenue on the upfront sale of our infrastructure and, if certain criteria are met, we will participate in revenue share generated by the AI clouds from their third-party customers.
We signed data center lease agreements with terms of ~15 years that are expected to commence between FY2028 and FY2029. We expect to reassign these data center leases to third parties.
We entered into guarantees to provide credit support on the land, power, and shell buildout to secure approximately 4.25 GW at SB Energy's PORTS-Pike Technology Campus in Ohio, which will exclusively host several generations of NVIDIA infrastructure under 20-year leases to OpenAI, subject to limited exceptions.

**Layout**

Single full-bleed black slide with the title centred in a band across the top (about 10 to 15 percent of the height) spanning nearly the full width. Three white-boxed financial tables are stacked in a left column (x about 4.8 to 47.3 percent) from top to bottom, and four green arrows lead from each table to a right-hand column of grey body paragraphs starting at x 56.4 percent. The NVIDIA logo sits at bottom right, and there is no visible slide number or footer.

**Visuals**

Four green line arrows, each with a filled triangular head pointing right toward the text. Arrow 1 is horizontal, pointing from the first table to paragraph 1. Arrow 2 is diagonal, rising from the bottom of the second table to paragraph 2. Arrow 3 is a shallow diagonal descending from the second table to paragraph 3. Arrow 4 is a shallow diagonal descending from the third table to paragraph 4. Table 1 (top, white box): caption 'Future commitments by fiscal year as of July 26, 2026, were as follows:'; columns Remainder of 2027, 2028, 2029, 2030, 2031, 2032 and thereafter, Total, with the header line '(In billions)'. Rows: Supply and capacity $92, $87, $88, $6, $5, $1, $279; Cloud service agreements 3, 8, 7, 6, 4, 1, 29; Data center leases not commenced —, 1, 1, 2, 1, 20, 25; Equity investments 18, 3, 2, 2, —, —, 25; Capital expenditures 7, 1, —, —, —, —, 8; Total $120, $100, $98, $16, $10, $22, $366. Table 2 (middle, white box): same caption and columns. Rows: AI cloud agreements $—, $6, $8, $7, $6, $9, $36; Data center leases not commenced for third party —, —, 1, 1, 1, 17, 20; Total $—, $6, $9, $8, $7, $26, $56. Table 3 (bottom, white box): caption 'The following table summarizes the maximum gross exposure related to our guarantees, including the SB Energy Corp. guarantees signed in August 2026 (in billions):'; rows Land, power, and shell guarantees for AI clouds $3.5; SB Energy Corp. guarantees 105.0; Total $108.5 (total ruled above). Table cell text is black on white with thin black rules under column headers and above totals; light grey row banding alternates on rows.

**Design**

Background: Solid black #000000 full-bleed.

**Elements and positions** (back to front; x, y, width, height as % of the slide; type, colour and alignment in brackets)

- shape at x 0, y 0, w 100, h 100 [#000000]: Full-slide black background fill (covers the entire slide, no border)
- table at x 4.8, y 21.8, w 42.5, h 25 [#FFFFFF]: Table 1: future commitments by fiscal year (in billions), supply and capacity, cloud service agreements, data center leases, equity investments, capital expenditures, total $366 (white rectangular image box, sharp corners, no shadow, placed image 1780 by 593 pixels)
- table at x 4.8, y 49.9, w 42.5, h 20.4 [#FFFFFF]: Table 2: AI cloud agreements and data center leases not commenced for third party, total $56 (white rectangular image box, sharp corners, no shadow, placed image 1787 by 484 pixels)
- table at x 4.8, y 73.5, w 42.5, h 12.8 [#FFFFFF]: Table 3: maximum gross exposure guarantees, total $108.5 (white rectangular image box, sharp corners, no shadow, placed image 1772 by 302 pixels)
- title at x 4.3, y 10.5, w 91.4, h 4.9 [NVIDIASans Medium, sans-serif, regular weight, normal case; font size 4.94% of slide height; #FFFFFF; center aligned]: NVIDIA Will Leverage Its Financial Capacity to Support the Ecosystem
- line at x 48, y 32.4, w 7.4, h 0.2 [#76B900]: Green arrow 1, horizontal, pointing to paragraph 1 (stroke about 2 px green, filled triangular arrowhead at right end)
- line at x 48, y 48.4, w 7.4, h 12.3 [#76B900]: Green arrow 2, diagonal rising, pointing to paragraph 2 (stroke about 2 px green, filled triangular arrowhead at upper-right end, starts at lower-left of table 2)
- line at x 48, y 66, w 7.4, h 2 [#76B900]: Green arrow 3, shallow diagonal descending, pointing to paragraph 3 (stroke about 2 px green, filled triangular arrowhead at lower-right end)
- line at x 48, y 82.2, w 7.4, h 1.8 [#76B900]: Green arrow 4, shallow diagonal descending, pointing to paragraph 4 (stroke about 2 px green, filled triangular arrowhead at lower-right end)
- body at x 56.4, y 28.2, w 38.5, h 8.4 [NVIDIASans Regular, sans-serif, regular weight; font size 1.98% of slide height; #A1A1A1; left aligned]: We’ve partnered with our extensive network of suppliers to secure the / critical components needed to meet demand for the next several years. / Our commitments increased from $119 billion last quarter to $279 billion, / primarily related to the procurement of memory. (line pitch about 2.1 percent of slide height)
- body at x 56.4, y 41.9, w 39.6, h 12.6 [NVIDIASans Regular, sans-serif, regular weight; font size 1.98% of slide height; #A1A1A1; left aligned]: We’ve partnered with leading AI clouds to enable broader access to our AI / infrastructure to serve AI startups, model builders, enterprises, research / organizations and sovereign customers. Under these agreements, we will / earn revenue on the upfront sale of our infrastructure and, if certain criteria / are met, we will participate in revenue share generated by the AI clouds / from their third-party customers. (line pitch about 2.1 percent of slide height)
- body at x 56.4, y 64.7, w 38.6, h 6.3 [NVIDIASans Regular, sans-serif, regular weight; font size 1.98% of slide height; #A1A1A1; left aligned]: We signed data center lease agreements with terms of ~15 years that are / expected to commence between FY2028 and FY2029. We expect to / reassign these data center leases to third parties. (line pitch about 2.1 percent of slide height)
- body at x 56.4, y 79.1, w 38.3, h 10.5 [NVIDIASans Regular, sans-serif, regular weight; font size 1.98% of slide height; #A1A1A1; left aligned]: We entered into guarantees to provide credit support on the land, power, / and shell buildout to secure approximately 4.25 GW at SB Energy's / PORTS-Pike Technology Campus in Ohio, which will exclusively host / several generations of NVIDIA infrastructure under 20-year leases to / OpenAI, subject to limited exceptions. (line pitch about 2.1 percent of slide height)
- logo at x 92.8, y 94.2, w 4.8, h 1.3 [#76B900]: NVIDIA logo (green eye mark with NVIDIA wordmark in white) (bottom-right corner, eye mark green, wordmark white)

**Measured from the file** (exact, read from the PDF rather than estimated; x, y, w, h as % of the slide; font size as % of the slide height)

- fill at x 0.0, y 0.0, w 100.0, h 100.0: #000000 (covers the slide)
- image at x 4.8, y 21.8, w 42.5, h 25.0: 1780 by 593 pixels
- image at x 4.8, y 49.9, w 42.5, h 20.4: 1787 by 484 pixels
- image at x 4.8, y 73.5, w 42.5, h 12.8: 1772 by 302 pixels
- text at x 4.3, y 10.5, w 91.4, h 4.9 [NVIDIASans-Medium; font size 4.94; #FFFFFF]: NVIDIA Will Leverage Its Financial Capacity to Support the Ecosystem
- text at x 56.4, y 28.2, w 36.8, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: We’ve partnered with our extensive network of suppliers to secure the
- text at x 56.4, y 30.4, w 37.3, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: critical components needed to meet demand for the next several years.
- text at x 56.4, y 32.5, w 38.5, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: Our commitments increased from $119 billion last quarter to $279 billion,
- text at x 56.4, y 34.6, w 25.3, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: primarily related to the procurement of memory.
- text at x 56.4, y 41.9, w 38.5, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: We’ve partnered with leading AI clouds to enable broader access to our AI
- text at x 56.4, y 44.0, w 37.7, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: infrastructure to serve AI startups, model builders, enterprises, research
- text at x 56.4, y 46.1, w 37.9, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: organizations and sovereign customers. Under these agreements, we will
- text at x 56.4, y 48.3, w 39.6, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: earn revenue on the upfront sale of our infrastructure and, if certain criteria
- text at x 56.4, y 50.4, w 37.0, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: are met, we will participate in revenue share generated by the AI clouds
- text at x 56.4, y 52.5, w 17.7, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: from their third-party customers.
- text at x 56.4, y 64.7, w 38.6, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: We signed data center lease agreements with terms of ~15 years that are
- text at x 56.4, y 66.8, w 35.4, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: expected to commence between FY2028 and FY2029. We expect to
- text at x 56.4, y 69.0, w 26.0, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: reassign these data center leases to third parties.
- text at x 56.4, y 79.1, w 38.3, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: We entered into guarantees to provide credit support on the land, power,
- text at x 56.4, y 81.2, w 34.6, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: and shell buildout to secure approximately 4.25 GW at SB Energy's
- text at x 56.4, y 83.4, w 35.0, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: PORTS-Pike Technology Campus in Ohio, which will exclusively host
- text at x 56.4, y 85.5, w 35.7, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: several generations of NVIDIA infrastructure under 20-year leases to
- text at x 56.4, y 87.6, w 19.9, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: OpenAI, subject to limited exceptions.

### Slide 10

**Review notes**

- Worth studying: Paired side-by-side curves of tokens per megawatt and cost per million tokens make the generational comparison readable at a glance.
- Review: The deck has a consistent black and NVIDIA green identity, action-style titles, and charts that carry real arguments, notably the paired curves on slide 10, the per-GW bars on slide 11 and the indexed networking chart on slide 17.
- Worth keeping: Keep the paired comparison charts on slide 10, the green-against-gray emphasis with annotated multiples, and the single-chart-with-numbered-list layout on slide 17.

**Text**

AI Factories Drive Token Economics — Compute is Revenue
Vera Rubin NVL72 delivers up to 30X better AI factory token throughput and 45X lower cost per million tokens.
DeepSeek-V4-Pro | SemiAnalysis AgentX
Tokens Per MW
80M
60M
40M
20M
0M
50
100
150
200
250
300
GB300 NVL72
Vera Rubin NVL72
30X
Interactivity (Tokens/Sec/User)
Dollars Per Million Tokens
$0.40
$0.30
$0.20
$0.10
$0.00
50
100
150
200
250
300
GB300 NVL72
Vera Rubin NVL72
45X
Lower
Interactivity (Tokens/Sec/User)

**Layout**

Two equal side-by-side chart panels sit in a two-column grid below a centred three-line header (title, subtitle, source line), with the NVIDIA logo bottom-right. The title is centred at the top of the slide, the left panel spans roughly x 5% to 49% and the right panel x 51% to 95%, both at y 31% to 86%, leaving equal side margins. No slide number or footer text is shown.

**Visuals**

Two line charts in outlined rounded-corner boxes. Left chart: title 'Tokens Per MW'; y-axis 0M to 80M in 20M steps, unlabelled units beyond the 'M' suffix; x-axis 'Interactivity (Tokens/Sec/User)' ticks 50, 100, 150, 200, 250, 300. Series 1 'Vera Rubin NVL72' (green #76B900) starts about x=84 at about 71M and falls smoothly to about 0M near x=272. Series 2 'GB300 NVL72' (grey #7F7F7F) starts about x=50 at about 58M, falls steeply to about 0M by x=170. A green dashed vertical double-headed arrow at x=150 spans from the GB300 curve to the Vera Rubin curve, labelled '30X' in green. Right chart: title 'Dollars Per Million Tokens'; y-axis $0.00 to $0.40 in $0.10 steps; x-axis 'Interactivity (Tokens/Sec/User)' ticks 50 to 300. Series 'GB300 NVL72' (grey) stays near $0.00 to about x=100, then rises steeply, exiting the top of the plot at about x=142 at $0.40. Series 'Vera Rubin NVL72' (green) stays near $0.00 until about x=190, then rises, exiting the top at about x=254 at $0.40. A green dashed vertical arrow at x=150 points down from the top of the plot to the baseline, labelled '45X Lower' in green. No legend box; series are identified by direct labels next to the curves. No values are printed on the curves; the 30X and 45X labels are the only printed values.

**Design**

Background: Solid black #000000 covering the whole slide.

**Elements and positions** (back to front; x, y, width, height as % of the slide; type, colour and alignment in brackets)

- shape at x 0, y 0, w 100, h 100 [#000000]: Slide background fill (Full-bleed solid fill, no gradient or texture.)
- shape at x 4.95, y 31, w 44.3, h 54.8 [#000000]: Left chart panel outline (Transparent fill, 1 px stroke #5A5A5A, corner radius about 1% of slide width, no shadow.)
- shape at x 50.7, y 31, w 44.3, h 54.8 [#000000]: Right chart panel outline (Transparent fill, 1 px stroke #5A5A5A, corner radius about 1% of slide width, no shadow.)
- chart at x 11, y 39.1, w 34.5, h 37.3 [#76B900]: Tokens Per MW line chart: Vera Rubin NVL72 and GB300 NVL72 curves with 30X arrow (Plot baseline at 0M in grey #7F7F7F, 1 px; no gridlines; green curve 2 px stroke #76B900; grey curve 2 px stroke #7F7F7F; dashed green arrow 1 px with arrowheads at both ends.)
- chart at x 57.5, y 39.1, w 34, h 37.3 [#76B900]: Dollars Per Million Tokens line chart: GB300 NVL72 and Vera Rubin NVL72 curves with 45X Lower arrow (Baseline grey #7F7F7F 1 px; no gridlines; grey curve 2 px #7F7F7F; green curve 2 px #76B900; dashed green arrow 1 px pointing down with arrowhead at the baseline.)
- title at x 11.4, y 10.5, w 77.2, h 4.9 [NVIDIASans Medium, sans-serif, normal weight, no caps, normal letter-spacing; font size 4.94% of slide height; #FFFFFF; center aligned]: AI Factories Drive Token Economics — Compute is Revenue
- subtitle at x 21.3, y 18.1, w 57.4, h 2 [NVIDIASans Regular, sans-serif, normal weight; font size 1.98% of slide height; #FFFFFF; center aligned]: Vera Rubin NVL72 delivers up to 30X better AI factory token throughput and 45X lower cost per million tokens.
- caption at x 39.3, y 22.2, w 21.5, h 2 [NVIDIASans Regular, sans-serif, normal weight; font size 1.98% of slide height; #A1A1A1; center aligned]: DeepSeek-V4-Pro | SemiAnalysis AgentX
- body at x 23, y 34.3, w 8.3, h 2 [NVIDIASans Medium, sans-serif, normal weight; font size 1.98% of slide height; #FFFFFF; center aligned]: Tokens Per MW
- body at x 66.1, y 34.3, w 13.6, h 2 [NVIDIASans Medium, sans-serif, normal weight; font size 1.98% of slide height; #FFFFFF; center aligned]: Dollars Per Million Tokens
- body at x 8, y 38.2, w 2.1, h 1.7 [NVIDIASans Regular, sans-serif, normal weight; font size 1.73% of slide height; #FFFFFF; right aligned]: 80M
- body at x 54.1, y 38.3, w 2.6, h 1.7 [NVIDIASans Regular, sans-serif, normal weight; font size 1.73% of slide height; #FFFFFF; right aligned]: $0.40
- body at x 62.9, y 39.8, w 6.4, h 1.7 [NVIDIASans Regular, sans-serif, normal weight; font size 1.73% of slide height; #7F7F7F; right aligned]: GB300 NVL72
- body at x 75.9, y 39.8, w 8.1, h 1.7 [NVIDIASans Regular, sans-serif, normal weight; font size 1.73% of slide height; #76B900; left aligned]: Vera Rubin NVL72
- body at x 69.3, y 56.5, w 3.6, h 4.9 [NVIDIASans Medium, sans-serif, normal weight; font size 2.22% of slide height; #76B900; center aligned]: 45X / Lower
- body at x 8, y 47.6, w 2.1, h 1.7 [NVIDIASans Regular, sans-serif, normal weight; font size 1.73% of slide height; #FFFFFF; right aligned]: 60M
- body at x 54.1, y 47.7, w 2.6, h 1.7 [NVIDIASans Regular, sans-serif, normal weight; font size 1.73% of slide height; #FFFFFF; right aligned]: $0.30
- body at x 23.6, y 66.6, w 2.3, h 2.2 [NVIDIASans Medium, sans-serif, normal weight; font size 2.22% of slide height; #76B900; center aligned]: 30X
- body at x 8, y 57, w 2.1, h 1.7 [NVIDIASans Regular, sans-serif, normal weight; font size 1.73% of slide height; #FFFFFF; right aligned]: 40M
- body at x 54.1, y 57, w 2.6, h 1.7 [NVIDIASans Regular, sans-serif, normal weight; font size 1.73% of slide height; #FFFFFF; right aligned]: $0.20
- body at x 8, y 66.3, w 2.1, h 1.7 [NVIDIASans Regular, sans-serif, normal weight; font size 1.73% of slide height; #FFFFFF; right aligned]: 20M
- body at x 54.1, y 66.4, w 2.6, h 1.7 [NVIDIASans Regular, sans-serif, normal weight; font size 1.73% of slide height; #FFFFFF; right aligned]: $0.10
- body at x 26.5, y 72.8, w 6.4, h 1.7 [NVIDIASans Regular, sans-serif, normal weight; font size 1.73% of slide height; #7F7F7F; left aligned]: GB300 NVL72
- body at x 37.5, y 72.8, w 8.2, h 1.7 [NVIDIASans Regular, sans-serif, normal weight; font size 1.73% of slide height; #76B900; right aligned]: Vera Rubin NVL72
- body at x 8.6, y 75.7, w 1.5, h 1.7 [NVIDIASans Regular, sans-serif, normal weight; font size 1.73% of slide height; #FFFFFF; right aligned]: 0M
- body at x 54.1, y 75.7, w 2.6, h 1.7 [NVIDIASans Regular, sans-serif, normal weight; font size 1.73% of slide height; #FFFFFF; right aligned]: $0.00
- body at x 10.4, y 78, w 1.2, h 1.7 [NVIDIASans Regular, sans-serif, normal weight; font size 1.73% of slide height; #FFFFFF; center aligned]: 50
- body at x 17, y 78, w 1.7, h 1.7 [NVIDIASans Regular, sans-serif, normal weight; font size 1.73% of slide height; #FFFFFF; center aligned]: 100
- body at x 23.9, y 78, w 1.7, h 1.7 [NVIDIASans Regular, sans-serif, normal weight; font size 1.73% of slide height; #FFFFFF; center aligned]: 150
- body at x 30.9, y 78, w 1.7, h 1.7 [NVIDIASans Regular, sans-serif, normal weight; font size 1.73% of slide height; #FFFFFF; center aligned]: 200
- body at x 37.8, y 78, w 1.7, h 1.7 [NVIDIASans Regular, sans-serif, normal weight; font size 1.73% of slide height; #FFFFFF; center aligned]: 250
- body at x 44.7, y 78, w 1.7, h 1.7 [NVIDIASans Regular, sans-serif, normal weight; font size 1.73% of slide height; #FFFFFF; center aligned]: 300
- body at x 57, y 78.1, w 1.2, h 1.7 [NVIDIASans Regular, sans-serif, normal weight; font size 1.73% of slide height; #FFFFFF; center aligned]: 50
- body at x 63.5, y 78.1, w 1.7, h 1.7 [NVIDIASans Regular, sans-serif, normal weight; font size 1.73% of slide height; #FFFFFF; center aligned]: 100
- body at x 70.3, y 78.1, w 1.7, h 1.7 [NVIDIASans Regular, sans-serif, normal weight; font size 1.73% of slide height; #FFFFFF; center aligned]: 150
- body at x 77, y 78.1, w 1.7, h 1.7 [NVIDIASans Regular, sans-serif, normal weight; font size 1.73% of slide height; #FFFFFF; center aligned]: 200
- body at x 83.8, y 78.1, w 1.7, h 1.7 [NVIDIASans Regular, sans-serif, normal weight; font size 1.73% of slide height; #FFFFFF; center aligned]: 250
- body at x 90.6, y 78.1, w 1.7, h 1.7 [NVIDIASans Regular, sans-serif, normal weight; font size 1.73% of slide height; #FFFFFF; center aligned]: 300
- caption at x 21.1, y 81.8, w 12, h 1.5 [NVIDIASans Regular, sans-serif, normal weight; font size 1.48% of slide height; #FFFFFF; center aligned]: Interactivity (Tokens/Sec/User)
- caption at x 66.9, y 81.8, w 12, h 1.5 [NVIDIASans Regular, sans-serif, normal weight; font size 1.48% of slide height; #FFFFFF; center aligned]: Interactivity (Tokens/Sec/User)
- logo at x 92.9, y 94.1, w 4.6, h 1.6 [#76B900]: NVIDIA logo (green eye mark with NVIDIA wordmark) (Wordmark in white #FFFFFF beside green #76B900 eye mark, bottom-right corner.)

**Measured from the file** (exact, read from the PDF rather than estimated; x, y, w, h as % of the slide; font size as % of the slide height)

- fill at x 0.0, y 0.0, w 100.0, h 100.0: #000000 (covers the slide)
- text at x 11.4, y 10.5, w 77.2, h 4.9 [NVIDIASans-Medium; font size 4.94; #FFFFFF]: AI Factories Drive Token Economics — Compute is Revenue
- text at x 21.3, y 18.1, w 57.4, h 2.0 [NVIDIASans-Regular; font size 1.98; #FFFFFF]: Vera Rubin NVL72 delivers up to 30X better AI factory token throughput and 45X lower cost per million tokens.
- text at x 39.3, y 22.2, w 21.5, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: DeepSeek-V4-Pro | SemiAnalysis AgentX
- text at x 23.0, y 34.3, w 8.3, h 2.0 [NVIDIASans-Medium; font size 1.98; #FFFFFF]: Tokens Per MW
- text at x 66.1, y 34.3, w 13.6, h 2.0 [NVIDIASans-Medium; font size 1.98; #FFFFFF]: Dollars Per Million Tokens
- text at x 8.0, y 38.2, w 2.1, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: 80M
- text at x 54.1, y 38.3, w 2.6, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: $0.40
- text at x 62.9, y 39.8, w 6.4, h 1.7 [NVIDIASans-Regular; font size 1.73; #7F7F7F]: GB300 NVL72
- text at x 75.9, y 39.8, w 8.1, h 1.7 [NVIDIASans-Regular; font size 1.73; #76B900]: Vera Rubin NVL72
- text at x 69.9, y 56.5, w 2.3, h 2.2 [NVIDIASans-Medium; font size 2.22; #76B900]: 45X
- text at x 69.3, y 59.2, w 3.6, h 2.2 [NVIDIASans-Medium; font size 2.22; #76B900]: Lower
- text at x 8.0, y 47.6, w 2.1, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: 60M
- text at x 54.1, y 47.7, w 2.6, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: $0.30
- text at x 23.6, y 66.6, w 2.3, h 2.2 [NVIDIASans-Medium; font size 2.22; #76B900]: 30X
- text at x 8.0, y 57.0, w 2.1, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: 40M
- text at x 54.1, y 57.0, w 2.6, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: $0.20
- text at x 8.0, y 66.3, w 2.1, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: 20M
- text at x 54.1, y 66.4, w 2.6, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: $0.10
- text at x 26.5, y 72.8, w 6.4, h 1.7 [NVIDIASans-Regular; font size 1.73; #7F7F7F]: GB300 NVL72
- text at x 37.5, y 72.8, w 8.2, h 1.7 [NVIDIASans-Regular; font size 1.73; #76B900]: Vera Rubin NVL72
- text at x 8.6, y 75.7, w 1.5, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: 0M
- text at x 54.1, y 75.7, w 2.6, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: $0.00
- text at x 10.4, y 78.0, w 1.2, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: 50
- text at x 17.0, y 78.0, w 1.7, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: 100
- text at x 23.9, y 78.0, w 1.7, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: 150
- text at x 30.9, y 78.0, w 1.7, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: 200
- text at x 37.8, y 78.0, w 1.7, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: 250
- text at x 44.7, y 78.0, w 1.7, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: 300
- text at x 57.0, y 78.1, w 1.2, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: 50
- text at x 63.5, y 78.1, w 1.7, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: 100
- text at x 70.3, y 78.1, w 1.7, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: 150
- text at x 77.0, y 78.1, w 1.7, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: 200
- text at x 83.8, y 78.1, w 1.7, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: 250
- text at x 90.6, y 78.1, w 1.7, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: 300
- text at x 21.1, y 81.8, w 12.0, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Interactivity (Tokens/Sec/User)
- text at x 66.9, y 81.8, w 12.0, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Interactivity (Tokens/Sec/User)

### Slide 11

**Review notes**

- Worth studying: A single trend chart of IT capex share across top cloud providers is paired with a clear statement of revenue opportunity per gigawatt, so the chart carries the argument directly.
- Review: The deck has a consistent black and NVIDIA green identity, action-style titles, and charts that carry real arguments, notably the paired curves on slide 10, the per-GW bars on slide 11 and the indexed networking chart on slide 17.
- What would change the score: I would rebuild slide 28 as a visual summary tied to the three growth pillars, and cut the intro paragraphs above the charts on slides 11 and 15 to one line each.

**Text**

NVIDIA Expanding AI Market While
Expanding Revenue Opportunity Per GW
NVIDIA continues to increase its share within the overall AI infrastructure market. NVIDIA’s per gigawatt revenue opportunity – based on its reference design – has expanded
from $18 billion during the Hopper generation to $25 billion with Blackwell and now $40 billion with Rubin. The growth in NVIDIA’s SAM (served available market) has been driven
by advancements in the GPU as well as a broadening in the company’s portfolio – which now includes stand-alone CPU racks for agentic AI, LPU racks for low-latency,
large-context inference, and scale-in networking for secure, performant AI factories. This is all a reflection of NVIDIA’s strategy in becoming a full-stack AI factory platform.
NVIDIA’s IT Capex Share Across the Top 5 CSPs
(Trailing 12-Month Basis)
50%
45%
40%
35%
30%
25%
20%
15%
10%
5%
0%
4Q22
1Q23
2Q23
3Q23
4Q23
1Q24
2Q24
3Q24
4Q24
1Q25
2Q25
3Q25
4Q25
1Q26
2Q26
CY27E
11%
11%
15%
20%
29%
35%
35%
36%
36%
35%
37%
39%
39%
40%
38%
44%
NVIDIA Increasing Platform Content Expands
Revenue Opportunity Per GW
Billions ($)
$60
$50
$40
$30
$20
$10
$0
Hopper
Blackwell
Rubin
Feynman

**Layout**

Two equal bordered panels sit side by side in the lower half, the left panel holding a line chart and the right panel a bar chart, with a centred two-line title across the top and a centred four-line grey paragraph between title and panels. The NVIDIA logo sits bottom right; the slide has no page number or footer text, and margins are roughly 5% on the left and right sides.

**Visuals**

Two charts. Left: a line chart titled NVIDIA’s IT Capex Share Across the Top 5 CSPs with subtitle (Trailing 12-Month Basis); y-axis 0% to 50% in 5% steps, no gridlines, x-axis categories 4Q22, 1Q23, 2Q23, 3Q23, 4Q23, 1Q24, 2Q24, 3Q24, 4Q24, 1Q25, 2Q25, 3Q25, 4Q25, 1Q26, 2Q26 (solid) then CY27E (dashed projection), labels rotated about 45 degrees. One green (#76B900) series with labelled values: 4Q22 11%, 1Q23 11%, 2Q23 15%, 3Q23 20%, 4Q23 29%, 1Q24 35%, 2Q24 35%, 3Q24 36%, 4Q24 36%, 1Q25 35%, 2Q25 37%, 3Q25 39%, 4Q25 39%, 1Q26 40%, 2Q26 38%; dashed extension to CY27E 44%. Data labels in small white text with thin leader ticks; no legend; no highlighted points. Right: a column chart titled NVIDIA Increasing Platform Content Expands / Revenue Opportunity Per GW with unit line Billions ($); y-axis $0 to $60 in $10 steps; categories Hopper, Blackwell, Rubin, Feynman; no data labels on bars. Hopper bar about $18 (from text), Blackwell $25 (from text), Rubin $40 (from text), Feynman about $60 by axis reading (no value printed). Bars are solid #76B900 except Feynman, which is a vertical gradient from near-black olive at top to #76B900 at bottom. No legend. Logo: NVIDIA eye mark in green with white NVIDIA wordmark, bottom right.

**Design**

Background: Solid black #000000 across the entire slide. Pure black background with white titles, grey body paragraph, white chart text and NVIDIA green accents. Typeface is NVIDIA Sans (a geometric humanist sans-serif): Medium weight for title and panel headings, Regular for body and axis labels. Title is about 4.9% of slide height (cap height about 3.5%), body paragraph about 2.0%, panel headings about 2.0%, axis labels about 1.7%, data labels about 1.2%. Title colour #FFFFFF, body colour #A1A1A1, accent #76B900. Panels have a thin grey outline about #4D4D4D, roughly 0.75 pt, with small corner radius, no fill and no shadow. Line is solid green about 3 pt with rounded joins; projection segment dashed green. Bars are flat with no corner radius and no shadow.

**Elements and positions** (back to front; x, y, width, height as % of the slide; type, colour and alignment in brackets)

- shape at x 0, y 0, w 100, h 100 [#000000]: Full-slide background fill (Covers the whole slide, behind everything.)
- title at x 23.7, y 5.2, w 52.9, h 9.2 [NVIDIASans Medium, sentence case; font size 4.94% of slide height; #FFFFFF; center aligned]: NVIDIA Expanding AI Market While / Expanding Revenue Opportunity Per GW (Two lines, line spacing about 1.1, centred.)
- body at x 4.4, y 18.1, w 91.6, h 10.6 [NVIDIASans Regular, sentence case; font size 1.98% of slide height; #A1A1A1; center aligned]: NVIDIA continues to increase its share within the overall AI infrastructure market. NVIDIA’s per gigawatt revenue opportunity – based on its reference design – has expanded / from $18 billion during the Hopper generation to $25 billion with Blackwell and now $40 billion with Rubin. The growth in NVIDIA’s SAM (served available market) has been driven / by advancements in the GPU as well as a broadening in the company’s portfolio – which now includes stand-alone CPU racks for agentic AI, LPU racks for low-latency, / large-context inference, and scale-in networking for secure, performant AI factories. This is all a reflection of NVIDIA’s strategy in becoming a full-stack AI factory platform. (Four centred lines, line spacing about 1.45.)
- shape at x 5, y 31, w 44.3, h 53.6 [#000000]: Left panel outline box (No fill, thin grey stroke about #4D4D4D 0.75 pt, corner radius about 0.5% of slide width, no shadow.)
- shape at x 50.7, y 31, w 44.3, h 53.6 [#000000]: Right panel outline box (No fill, thin grey stroke about #4D4D4D 0.75 pt, corner radius about 0.5% of slide width, no shadow.)
- body at x 14.7, y 34.3, w 25.2, h 2 [NVIDIASans Medium, sentence case; font size 1.98% of slide height; #FFFFFF; center aligned]: NVIDIA’s IT Capex Share Across the Top 5 CSPs (Panel heading.)
- caption at x 22.2, y 37, w 9.8, h 1.5 [NVIDIASans Regular; font size 1.48% of slide height; #FFFFFF; center aligned]: (Trailing 12-Month Basis)
- chart at x 12, y 40, w 37, h 40 [#76B900]: Line chart of NVIDIA IT capex share across top 5 CSPs, plotted in green #76B900 about 3 pt rounded, dashed from 2Q26 to CY27E (Plot baseline grey axis line at 0% from x about 11 to 46 percent of slide. Points about 4Q22 11%, 1Q23 11%, 2Q23 15%, 3Q23 20%, 4Q23 29%, 1Q24 35%, 2Q24 35%, 3Q24 36%, 4Q24 36%, 1Q25 35%, 2Q25 37%, 3Q25 39%, 4Q25 39%, 1Q26 40%, 2Q26 38%, dashed to CY27E 44%. Labels in small white with short grey leader lines.)
- body at x 8.2, y 41.4, w 2, h 1.7 [NVIDIASans Regular; font size 1.73% of slide height; #FFFFFF; right aligned]: 50%
- body at x 8.2, y 44.8, w 2, h 1.7 [NVIDIASans Regular; font size 1.73% of slide height; #FFFFFF; right aligned]: 45%
- body at x 8.2, y 48.3, w 2, h 1.7 [NVIDIASans Regular; font size 1.73% of slide height; #FFFFFF; right aligned]: 40%
- body at x 8.2, y 51.7, w 2, h 1.7 [NVIDIASans Regular; font size 1.73% of slide height; #FFFFFF; right aligned]: 35%
- body at x 8.2, y 55.1, w 2, h 1.7 [NVIDIASans Regular; font size 1.73% of slide height; #FFFFFF; right aligned]: 30%
- body at x 8.2, y 58.5, w 2, h 1.7 [NVIDIASans Regular; font size 1.73% of slide height; #FFFFFF; right aligned]: 25%
- body at x 8.2, y 62, w 2, h 1.7 [NVIDIASans Regular; font size 1.73% of slide height; #FFFFFF; right aligned]: 20%
- body at x 8.2, y 65.4, w 2, h 1.7 [NVIDIASans Regular; font size 1.73% of slide height; #FFFFFF; right aligned]: 15%
- body at x 8.2, y 68.8, w 2, h 1.7 [NVIDIASans Regular; font size 1.73% of slide height; #FFFFFF; right aligned]: 10%
- body at x 8.7, y 72.2, w 1.4, h 1.7 [NVIDIASans Regular; font size 1.73% of slide height; #FFFFFF; right aligned]: 5%
- body at x 8.7, y 75.7, w 1.4, h 1.7 [NVIDIASans Regular; font size 1.73% of slide height; #FFFFFF; right aligned]: 0%
- body at x 20, y 78, w 3, h 2.5 [NVIDIASans Regular; font size 1.73% of slide height; #FFFFFF; right aligned]: 4Q22 (Rotated about 45 degrees.)
- body at x 23, y 78, w 3, h 2.5 [NVIDIASans Regular; font size 1.73% of slide height; #FFFFFF; right aligned]: 1Q23 (Rotated about 45 degrees.)
- body at x 26, y 78, w 3, h 2.5 [NVIDIASans Regular; font size 1.73% of slide height; #FFFFFF; right aligned]: 2Q23 (Rotated about 45 degrees.)
- body at x 29, y 78, w 3, h 2.5 [NVIDIASans Regular; font size 1.73% of slide height; #FFFFFF; right aligned]: 3Q23 (Rotated about 45 degrees.)
- body at x 32, y 78, w 3, h 2.5 [NVIDIASans Regular; font size 1.73% of slide height; #FFFFFF; right aligned]: 4Q23 (Rotated about 45 degrees.)
- body at x 35, y 78, w 3, h 2.5 [NVIDIASans Regular; font size 1.73% of slide height; #FFFFFF; right aligned]: 1Q24 (Rotated about 45 degrees.)
- body at x 38, y 78, w 3, h 2.5 [NVIDIASans Regular; font size 1.73% of slide height; #FFFFFF; right aligned]: 2Q24 (Rotated about 45 degrees.)
- body at x 41, y 78, w 3, h 2.5 [NVIDIASans Regular; font size 1.73% of slide height; #FFFFFF; right aligned]: 3Q24 (Rotated about 45 degrees.)
- body at x 44, y 78, w 3, h 2.5 [NVIDIASans Regular; font size 1.73% of slide height; #FFFFFF; right aligned]: 4Q24 (Rotated about 45 degrees.)
- body at x 47, y 78, w 3, h 2.5 [NVIDIASans Regular; font size 1.73% of slide height; #FFFFFF; right aligned]: 1Q25 (Rotated about 45 degrees.)
- body at x 50, y 78, w 3, h 2.5 [NVIDIASans Regular; font size 1.73% of slide height; #FFFFFF; right aligned]: 2Q25 (Rotated about 45 degrees.)
- body at x 53, y 78, w 3, h 2.5 [NVIDIASans Regular; font size 1.73% of slide height; #FFFFFF; right aligned]: 3Q25 (Rotated about 45 degrees.)
- body at x 56, y 78, w 3, h 2.5 [NVIDIASans Regular; font size 1.73% of slide height; #FFFFFF; right aligned]: 4Q25 (Rotated about 45 degrees.)
- body at x 59, y 78, w 3, h 2.5 [NVIDIASans Regular; font size 1.73% of slide height; #FFFFFF; right aligned]: 1Q26 (Rotated about 45 degrees.)
- body at x 62, y 78, w 3, h 2.5 [NVIDIASans Regular; font size 1.73% of slide height; #FFFFFF; right aligned]: 2Q26 (Rotated about 45 degrees.)
- body at x 84, y 78, w 4, h 2.5 [NVIDIASans Regular; font size 1.73% of slide height; #FFFFFF; right aligned]: CY27E (Rotated about 45 degrees.)
- body at x 10.9, y 66.4, w 1.4, h 1.2 [NVIDIASans Regular; font size 1.24% of slide height; #FFFFFF; center aligned]: 11%
- body at x 12.5, y 64.8, w 1.4, h 1.2 [NVIDIASans Regular; font size 1.24% of slide height; #FFFFFF; center aligned]: 11%
- body at x 13.9, y 63.2, w 1.4, h 1.2 [NVIDIASans Regular; font size 1.24% of slide height; #FFFFFF; center aligned]: 15%
- body at x 15, y 59.8, w 1.4, h 1.2 [NVIDIASans Regular; font size 1.24% of slide height; #FFFFFF; center aligned]: 20%
- body at x 16.2, y 54.9, w 1.4, h 1.2 [NVIDIASans Regular; font size 1.24% of slide height; #FFFFFF; center aligned]: 29%
- body at x 17.7, y 50.9, w 1.4, h 1.2 [NVIDIASans Regular; font size 1.24% of slide height; #FFFFFF; center aligned]: 35%
- body at x 19.7, y 49.6, w 1.4, h 1.2 [NVIDIASans Regular; font size 1.24% of slide height; #FFFFFF; center aligned]: 35%
- body at x 21.7, y 48.6, w 3.6, h 1.3 [NVIDIASans Regular; font size 1.24% of slide height; #FFFFFF; center aligned]: 36% 36%
- body at x 25.6, y 48.1, w 1.4, h 1.2 [NVIDIASans Regular; font size 1.24% of slide height; #FFFFFF; center aligned]: 35%
- body at x 27.2, y 47.7, w 1.4, h 1.2 [NVIDIASans Regular; font size 1.24% of slide height; #FFFFFF; center aligned]: 37%
- body at x 29.1, y 46.1, w 5.2, h 1.7 [NVIDIASans Regular; font size 1.24% of slide height; #FFFFFF; center aligned]: 39% 39% 40%
- body at x 34.8, y 47.3, w 1.4, h 1.2 [NVIDIASans Regular; font size 1.24% of slide height; #FFFFFF; center aligned]: 38%
- body at x 44.4, y 44, w 1.4, h 1.2 [NVIDIASans Regular; font size 1.24% of slide height; #FFFFFF; center aligned]: 44%
- body at x 61, y 34.3, w 24.1, h 4.2 [NVIDIASans Medium, sentence case; font size 1.98% of slide height; #FFFFFF; center aligned]: NVIDIA Increasing Platform Content Expands / Revenue Opportunity Per GW
- caption at x 70.9, y 39.1, w 4, h 1.5 [NVIDIASans Regular; font size 1.48% of slide height; #FFFFFF; center aligned]: Billions ($)
- body at x 54.5, y 41.5, w 1.7, h 1.7 [NVIDIASans Regular; font size 1.73% of slide height; #FFFFFF; right aligned]: $60
- body at x 54.5, y 47.2, w 1.7, h 1.7 [NVIDIASans Regular; font size 1.73% of slide height; #FFFFFF; right aligned]: $50
- body at x 54.5, y 52.9, w 1.7, h 1.7 [NVIDIASans Regular; font size 1.73% of slide height; #FFFFFF; right aligned]: $40
- body at x 54.5, y 58.6, w 1.7, h 1.7 [NVIDIASans Regular; font size 1.73% of slide height; #FFFFFF; right aligned]: $30
- body at x 54.5, y 64.3, w 1.7, h 1.7 [NVIDIASans Regular; font size 1.73% of slide height; #FFFFFF; right aligned]: $20
- body at x 54.5, y 70, w 1.7, h 1.7 [NVIDIASans Regular; font size 1.73% of slide height; #FFFFFF; right aligned]: $10
- body at x 55.1, y 75.7, w 1.2, h 1.7 [NVIDIASans Regular; font size 1.73% of slide height; #FFFFFF; right aligned]: $0
- shape at x 60.1, y 66.3, w 2.8, h 10.2 [#76B900]: Hopper bar (Flat solid bar, no corners, no shadow, value about $18 read from body text.)
- shape at x 68.7, y 62.3, w 2.8, h 14.3 [#76B900]: Blackwell bar (Flat solid bar, no corners, no shadow, value $25 read from body text.)
- shape at x 77.2, y 53.8, w 2.8, h 22.8 [#76B900]: Rubin bar (Flat solid bar, no corners, no shadow, value $40 read from body text.)
- shape at x 85.8, y 42.3, w 2.8, h 34.3 [#76B900]: Feynman bar, vertical gradient (Gradient from near-black olive #1A2400 at top to #76B900 at bottom, 82 by 556 pixels, about $60 by axis reading, no value printed.)
- body at x 59.7, y 78, w 3.4, h 1.7 [NVIDIASans Regular; font size 1.73% of slide height; #FFFFFF; center aligned]: Hopper
- body at x 68, y 78, w 4.2, h 1.7 [NVIDIASans Regular; font size 1.73% of slide height; #FFFFFF; center aligned]: Blackwell
- body at x 77.3, y 78, w 2.6, h 1.7 [NVIDIASans Regular; font size 1.73% of slide height; #FFFFFF; center aligned]: Rubin
- body at x 85.1, y 78, w 4.2, h 1.7 [NVIDIASans Regular; font size 1.73% of slide height; #FFFFFF; center aligned]: Feynman
- line at x 57.2, y 75.9, w 28.3, h 0.1 [#FFFFFF]: Bar chart baseline x-axis (Thin white or grey horizontal line at $0, about 0.5 pt.)
- logo at x 92.8, y 94, w 7, h 1.8 [#76B900]: NVIDIA logo: green eye mark with white NVIDIA wordmark, bottom right corner (Eye icon in #76B900, wordmark in #FFFFFF.)

**Measured from the file** (exact, read from the PDF rather than estimated; x, y, w, h as % of the slide; font size as % of the slide height)

- fill at x 0.0, y 0.0, w 100.0, h 100.0: #000000 (covers the slide)
- fill at x 68.7, y 62.3, w 2.8, h 14.3: #76B900
- fill at x 77.2, y 53.8, w 2.8, h 22.8: #76B900
- image at x 85.8, y 42.3, w 2.8, h 34.3: 82 by 556 pixels
- image at x 85.8, y 42.3, w 2.8, h 34.3: 82 by 556 pixels
- text at x 27.1, y 5.2, w 46.4, h 4.9 [NVIDIASans-Medium; font size 4.94; #FFFFFF]: NVIDIA Expanding AI Market While
- text at x 23.7, y 10.5, w 52.9, h 4.9 [NVIDIASans-Medium; font size 4.94; #FFFFFF]: Expanding Revenue Opportunity Per GW
- text at x 5.5, y 18.1, w 89.4, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: NVIDIA continues to increase its share within the overall AI infrastructure market. NVIDIA’s per gigawatt revenue opportunity – based on its reference design – has expanded
- text at x 4.4, y 20.7, w 91.6, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: from $18 billion during the Hopper generation to $25 billion with Blackwell and now $40 billion with Rubin. The growth in NVIDIA’s SAM (served available market) has been driven
- text at x 7.3, y 23.3, w 85.6, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: by advancements in the GPU as well as a broadening in the company’s portfolio – which now includes stand-alone CPU racks for agentic AI, LPU racks for low-latency,
- text at x 5.6, y 25.9, w 88.8, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: large-context inference, and scale-in networking for secure, performant AI factories. This is all a reflection of NVIDIA’s strategy in becoming a full-stack AI factory platform.
- text at x 14.7, y 34.3, w 25.2, h 2.0 [NVIDIASans-Medium; font size 1.98; #FFFFFF]: NVIDIA’s IT Capex Share Across the Top 5 CSPs
- text at x 61.0, y 34.3, w 24.1, h 2.0 [NVIDIASans-Medium; font size 1.98; #FFFFFF]: NVIDIA Increasing Platform Content Expands
- text at x 65.1, y 36.5, w 15.5, h 2.0 [NVIDIASans-Medium; font size 1.98; #FFFFFF]: Revenue Opportunity Per GW
- text at x 8.2, y 41.4, w 2.0, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: 50%
- text at x 54.5, y 41.5, w 1.7, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: $60
- text at x 22.2, y 37.0, w 9.8, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: (Trailing 12-Month Basis)
- text at x 8.2, y 44.8, w 2.0, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: 45%
- text at x 70.9, y 39.1, w 4.0, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Billions ($)
- text at x 54.5, y 47.2, w 1.7, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: $50
- text at x 8.2, y 48.3, w 2.0, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: 40%
- text at x 8.2, y 51.7, w 2.0, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: 35%
- text at x 54.5, y 52.9, w 1.7, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: $40
- text at x 8.2, y 55.1, w 2.0, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: 30%
- text at x 8.2, y 58.5, w 2.0, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: 25%
- text at x 54.5, y 58.6, w 1.7, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: $30
- text at x 44.4, y 44.0, w 1.4, h 1.2 [NVIDIASans-Regular; font size 1.24; #FFFFFF]: 44%
- text at x 8.2, y 62.0, w 2.0, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: 20%
- text at x 29.1, y 46.1, w 5.2, h 1.7 [NVIDIASans-Regular; font size 1.24; #FFFFFF]: 39% 39% 40%
- text at x 54.5, y 64.3, w 1.7, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: $20
- text at x 8.2, y 65.4, w 2.0, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: 15%
- text at x 27.2, y 47.7, w 1.4, h 1.2 [NVIDIASans-Regular; font size 1.24; #FFFFFF]: 37%
- text at x 34.8, y 47.3, w 1.4, h 1.2 [NVIDIASans-Regular; font size 1.24; #FFFFFF]: 38%
- text at x 25.6, y 48.1, w 1.4, h 1.2 [NVIDIASans-Regular; font size 1.24; #FFFFFF]: 35%
- text at x 21.7, y 48.6, w 3.6, h 1.3 [NVIDIASans-Regular; font size 1.24; #FFFFFF]: 36% 36%
- text at x 8.2, y 68.8, w 2.0, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: 10%
- text at x 19.7, y 49.6, w 1.4, h 1.2 [NVIDIASans-Regular; font size 1.24; #FFFFFF]: 35%
- text at x 54.5, y 70.0, w 1.7, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: $10
- text at x 17.7, y 50.9, w 1.4, h 1.2 [NVIDIASans-Regular; font size 1.24; #FFFFFF]: 35%
- text at x 8.7, y 72.2, w 1.4, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: 5%
- text at x 8.7, y 75.7, w 1.4, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: 0%
- text at x 55.1, y 75.7, w 1.2, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: $0
- text at x 16.2, y 54.9, w 1.4, h 1.2 [NVIDIASans-Regular; font size 1.24; #FFFFFF]: 29%
- text at x 59.7, y 78.0, w 3.4, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: Hopper
- text at x 68.0, y 78.0, w 4.2, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: Blackwell
- text at x 77.3, y 78.0, w 2.6, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: Rubin
- text at x 85.1, y 78.0, w 4.2, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: Feynman
- text at x 15.0, y 59.8, w 1.4, h 1.2 [NVIDIASans-Regular; font size 1.24; #FFFFFF]: 20%
- text at x 13.9, y 63.2, w 1.4, h 1.2 [NVIDIASans-Regular; font size 1.24; #FFFFFF]: 15%
- text at x 12.5, y 64.8, w 1.4, h 1.2 [NVIDIASans-Regular; font size 1.24; #FFFFFF]: 11%
- text at x 10.9, y 66.4, w 1.4, h 1.2 [NVIDIASans-Regular; font size 1.24; #FFFFFF]: 11%

### Slide 12

**Text**

Increasing Share of Data Center CapEx
Vera Rubin NVL72
Compute Node
Model Processing
Vera MGX
Compute Node
Secure Agent Processing
Vera CMX
Storage Node
Context KV$ Offload
Vera DPX
Storage Node
Data Processing

**Layout**

A single centred title sits across the top of the slide at about 10.5% from the top. Below it, four server node units are placed in a horizontal row across the width, each with a centred two-line heading and a grey descriptor above it. Their heights step up from left to right, with the Vera DPX label highest. The units are arranged in a loose row with no visible grid, rules, footer or slide number, and the margins are generous on all sides.

**Visuals**

Four photographic renderings of data centre server units, shown at three-quarter angles on a black reflective surface with soft mirror reflections beneath each unit. From left to right: (1) Vera Rubin NVL72 Compute Node, a flat tray-style board with gold-edged handles, several gold-lidded processor modules with green and blue centres, a central black heat spine and a dark board with small connectors; (2) Vera MGX Compute Node, a tray with six visible dark processor packages with blue and green centres, four small gold-framed modules along the bottom edge, and a silver-grey lower panel with gold handles; (3) Vera CMX Storage Node, an open chassis with a black interior board, two large processor packages with green centres, several smaller modules and connectors, and a silver lower panel with gold-coloured vents; (4) Vera DPX Storage Node, a taller rack-style chassis with gold-coloured heatsink fins on top, a dark circuit board with a row of connectors, and a gold-toned front grille with vent panels. Lighting is a single soft key light from above, producing highlights on the metal edges. Dominant colours are black, gold, silver-grey and small green and blue accents. Each unit is shown whole, not cropped, and is not boxed; the photograph covers the full slide.

**Design**

Background: Solid black #000000 background, with the four units standing on a dark reflective surface.

**Elements and positions** (back to front; x, y, width, height as % of the slide; type, colour and alignment in brackets)

- shape at x 0, y 0, w 100, h 100 [#000000]: Full-slide solid black fill (Covers the whole slide; no border or shadow.)
- photo at x 0, y 0, w 100, h 100 [#000000]: Full-slide background image of a black surface, which has no visible content apart from the reflective floor under the units (Full-bleed, uncropped, no overlay.)
- title at x 24.5, y 10.5, w 51, h 4.9 [NVIDIASans, sans-serif, Medium weight, normal case, no letter-spacing; font size 4.94% of slide height; #FFFFFF; center aligned]: Increasing Share of Data Center CapEx
- photo at x 27, y 46.5, w 40, h 50 [#A8925A]: Vera DPX Storage Node server unit, rack-style chassis with gold heatsink fins, dark circuit board and gold front grille, shown at a three-quarter angle with reflection below (Full-colour render on black background, not boxed, not cropped; reflection extends to the bottom edge.)
- photo at x 4.5, y 53.5, w 23.5, h 30 [#B5A36C]: Vera Rubin NVL72 Compute Node server tray with gold-lidded processor modules and a central black heat spine, shown at a three-quarter angle with reflection below (Full-colour render on black background, not boxed, not cropped; reflection below.)
- photo at x 29, y 53, w 18.5, h 31 [#B5A36C]: Vera MGX Compute Node server tray with six dark processor packages and gold handles, shown at a three-quarter angle with reflection below (Full-colour render on black background, not boxed, not cropped; reflection below.)
- photo at x 52, y 51.5, w 21, h 31.5 [#B5A36C]: Vera CMX Storage Node open chassis with black interior board and silver-grey lower panel, shown at a three-quarter angle with reflection below (Full-colour render on black background, not boxed, not cropped; reflection below.)
- title at x 76.2, y 36.1, w 5.8, h 2.2 [NVIDIASans, sans-serif, Medium weight, normal case; font size 2.22% of slide height; #FFFFFF; center aligned]: Vera DPX
- title at x 74.9, y 39, w 8.1, h 2.2 [NVIDIASans, sans-serif, Medium weight, normal case; font size 2.22% of slide height; #FFFFFF; center aligned]: Storage Node
- body at x 74.7, y 42.1, w 8.4, h 2 [NVIDIASans, sans-serif, Regular weight, normal case; font size 1.98% of slide height; #A1A1A1; center aligned]: Data Processing
- title at x 16.1, y 40.2, w 10.9, h 2.2 [NVIDIASans, sans-serif, Medium weight, normal case; font size 2.22% of slide height; #FFFFFF; center aligned]: Vera Rubin NVL72
- title at x 16.9, y 43.1, w 8.9, h 2.2 [NVIDIASans, sans-serif, Medium weight, normal case; font size 2.22% of slide height; #FFFFFF; center aligned]: Compute Node
- body at x 16.8, y 46.2, w 9.2, h 2 [NVIDIASans, sans-serif, Regular weight, normal case; font size 1.98% of slide height; #A1A1A1; center aligned]: Model Processing
- title at x 38, y 40.2, w 6.1, h 2.2 [NVIDIASans, sans-serif, Medium weight, normal case; font size 2.22% of slide height; #FFFFFF; center aligned]: Vera MGX
- title at x 36.4, y 43.1, w 8.9, h 2.2 [NVIDIASans, sans-serif, Medium weight, normal case; font size 2.22% of slide height; #FFFFFF; center aligned]: Compute Node
- body at x 34.4, y 46.2, w 12.9, h 2 [NVIDIASans, sans-serif, Regular weight, normal case; font size 1.98% of slide height; #A1A1A1; center aligned]: Secure Agent Processing
- title at x 56.2, y 40.2, w 5.8, h 2.2 [NVIDIASans, sans-serif, Medium weight, normal case; font size 2.22% of slide height; #FFFFFF; center aligned]: Vera CMX
- title at x 55.1, y 43.1, w 8.1, h 2.2 [NVIDIASans, sans-serif, Medium weight, normal case; font size 2.22% of slide height; #FFFFFF; center aligned]: Storage Node
- body at x 53.7, y 46.2, w 10.8, h 2 [NVIDIASans, sans-serif, Regular weight, normal case; font size 1.98% of slide height; #A1A1A1; center aligned]: Context KV$ Offload

**Measured from the file** (exact, read from the PDF rather than estimated; x, y, w, h as % of the slide; font size as % of the slide height)

- fill at x 0.0, y 0.0, w 100.0, h 100.0: #000000 (covers the slide)
- image at x 0.0, y 0.0, w 100.0, h 100.0: 7680 by 4320 pixels (covers the slide)
- text at x 24.5, y 10.5, w 51.0, h 4.9 [NVIDIASans-Medium; font size 4.94; #FFFFFF]: Increasing Share of Data Center CapEx
- text at x 76.2, y 36.1, w 5.8, h 2.2 [NVIDIASans-Medium; font size 2.22; #FFFFFF]: Vera DPX
- text at x 74.9, y 39.0, w 8.1, h 2.2 [NVIDIASans-Medium; font size 2.22; #FFFFFF]: Storage Node
- text at x 16.1, y 40.2, w 10.9, h 2.2 [NVIDIASans-Medium; font size 2.22; #FFFFFF]: Vera Rubin NVL72
- text at x 38.0, y 40.2, w 6.1, h 2.2 [NVIDIASans-Medium; font size 2.22; #FFFFFF]: Vera MGX
- text at x 56.2, y 40.2, w 5.8, h 2.2 [NVIDIASans-Medium; font size 2.22; #FFFFFF]: Vera CMX
- text at x 16.9, y 43.1, w 8.9, h 2.2 [NVIDIASans-Medium; font size 2.22; #FFFFFF]: Compute Node
- text at x 36.4, y 43.1, w 8.9, h 2.2 [NVIDIASans-Medium; font size 2.22; #FFFFFF]: Compute Node
- text at x 55.1, y 43.1, w 8.1, h 2.2 [NVIDIASans-Medium; font size 2.22; #FFFFFF]: Storage Node
- text at x 74.7, y 42.1, w 8.4, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: Data Processing
- text at x 16.8, y 46.2, w 9.2, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: Model Processing
- text at x 34.4, y 46.2, w 12.9, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: Secure Agent Processing
- text at x 53.7, y 46.2, w 10.8, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: Context KV$ Offload

### Slide 13

**Review notes**

- Review: Weaknesses are consequential: many slides depend on small gray body text and tiny labels (slides 4, 8, 13, 18, 21, 22), the financial tables on slide 9 are unreadable at presentation size, and slide 2 is a dense disclaimer wall.
- Suggested repair: Enlarge or simplify the small body text and diagram labels on slides 4, 8, 13, 18 and 21, and replace the screenshot-style tables on slide 9 with a few legible headline figures.

**Text**

Announcing NVIDIA Open Agent Security Platform
An Open Reference Design with NVIDIA OpenShell and NVIDIA Sentry
OpenShell is a policy-controlled runtime between an agent and the enterprise systems it can affect – data, files, credentials, tools, APIs, models, and networks.
OpenShell addresses the agent security problem: 1) Prevents uncontrolled access to enterprise data, 2) Limits credential exposure and misuse, 3) Controls agent tool use and code execution, 4) Applies default-deny network egress where appropriate, 5) Helps contain prompt injection, malicious retrieved content, and harmful tool calls, and 6) Creates an auditable runtime boundary for long-running autonomous workflows.
Secure Runtime Boundary
Governs and Mediates All Actions, Enforces Policy at Runtime
In-Silicon Out-of-Band Monitor
Agent Reasoning Inspection and Tamper-Proof Telemetry
NVIDIA OpenShell
Gateway
Supervisor
Policy Prover
Agent Harness
Agent Sandbox
NVIDIA Sentry
Hardware-Based Enforcement
Continuous Agent Monitoring and Threat Detection
Trusted Telemetry and Detection Pipeline
Millisecond-Scale Containment and Quarantine
Vera CPU
BlueField-4
Prompt / Tools / Data / Model / Policies
Prompt
Tools
Data
Model
Vera MGX Compute Node
Secure Agent Processing

**Layout**

Full-bleed black slide with a centred three-part composition: a title block across the top (main title, subtitle, and two paragraphs of grey body text centred), then three columns below. The left column holds the green-outlined NVIDIA OpenShell panel with a lock icon at its top-left corner; the centre column holds the tall Vera MGX compute node rack image with its caption beneath; the right column holds the green-outlined NVIDIA Sentry panel with a lock icon at its top-right corner and four bottom tiles beneath it. Each column has a centred heading and subheading above its panel. Margins are about 6 to 7 percent on the left and right. The NVIDIA logo sits bottom right. There is no visible slide number or footer.

**Visuals**

1) Photograph/illustration: a tall, dark, vertical compute-node rack centred horizontally at about x 40 to 60 percent and spanning roughly y 32 to 96 percent. It shows stacked black chassis with heatsink fins, four rows of square chips with coloured blue and pink-purple glowing dies on the upper and middle sections, two gold-brown chips near the bottom, a silver metal cover plate at the base, and gold-tan handles at the foot. Lighting is dim and directional with the chips lit. Dominant colours are black, dark grey, silver and gold-tan with small blue and magenta accents. The image is unboxed and sits directly on the black background; it is not cropped. Text labels Vera CPU and BlueField-4 sit on small dark boxes with thin white outlines over the chassis.
2) Diagram: a green (#76B900) double-headed arrow runs horizontally from the OpenShell panel to the rack at about y 50 percent, with arrowheads at both ends and a thin line. A green vertical double-headed arrow runs from the upper rack chips down to the bottom chips at about x 44 percent, with arrowheads at both ends. A green horizontal arrow at about y 76 percent runs from the rack's lower chips to the Sentry panel, with an arrowhead pointing right into the panel. Grey stacked labels Prompt, Tools, Data, Model, Policies sit between the rack and the Sentry panel at about x 60 percent.
3) Shapes: NVIDIA OpenShell and NVIDIA Sentry are rounded-corner rectangles with a 2 px green (#76B900) outline and no fill. Inside OpenShell are three dark grey tiles (Gateway, Supervisor, Policy Prover) and one larger dark grey tile (Agent Harness) inside a white dashed rounded rectangle labelled Agent Sandbox. Inside Sentry are four dark grey tiles in a vertical stack. Below Sentry are four dark grey tiles in a row, each linked by a grey dashed double-headed arrow to the Sentry panel.
4) Icons: two green outline padlocks, one at the top-left corner of the OpenShell panel and one at the top-right corner of the Sentry panel.
5) Logo: the NVIDIA logo in green and white at bottom right.

**Design**

Background: Solid black #000000 full-bleed background. Black (#000000) background with no gradient. Titles are white (#FFFFFF) NVIDIA Sans Medium, about 4.9 percent of slide height for the main title and about 2.2 percent for column headings. Subtitle is white NVIDIA Sans Medium at about 2.0 percent. Body paragraphs are grey (#A1A1A1) NVIDIA Sans Regular at about 2.0 percent. Column subheadings are white NVIDIA Sans Regular at about 1.7 percent. Panel headings NVIDIA OpenShell and NVIDIA Sentry are green (#76B900) NVIDIA Sans Bold at about 2.0 percent. Accent colour is NVIDIA green #76B900. Tile labels are white NVIDIA Sans Regular at about 1.7 percent, with the Agent Sandbox label in bold at about 1.5 percent. Source tile labels for Prompt, Data and Model are grey #757575 at about 1.7 percent. Tiles are flat dark grey #262626 rectangles with very slight corner rounding (about 1 percent of width) and no shadows. Panels have 2 px green outlines with square corners of small radius. Dashed sandbox outline is white, dashed, about 2 px. Arrows are green, about 2 px. Sans-serif throughout.

**Elements and positions** (back to front; x, y, width, height as % of the slide; type, colour and alignment in brackets)

- shape at x 0, y 0, w 100, h 100 [#000000]: full-slide black background (no border, behind all other elements)
- shape at x 7.2, y 46, w 26, h 33 [#000000]: NVIDIA OpenShell panel outline (no fill; 2 px stroke #76B900; corner radius about 0.5 percent of width; sits behind tiles and dashed sandbox)
- shape at x 66.8, y 46, w 26, h 33 [#000000]: NVIDIA Sentry panel outline (no fill; 2 px stroke #76B900; corner radius about 0.5 percent of width)
- shape at x 19.9, y 53.2, w 11.9, h 18.7 [#000000]: Agent Sandbox dashed rounded rectangle (no fill; white dashed stroke about 2 px; corner radius about 1 percent of width)
- icon at x 6.2, y 42.4, w 2.4, h 5.8 [#76B900]: green padlock at top-left of OpenShell panel (outline padlock, stroke about 3 px, overlaps panel top-left corner)
- icon at x 91.4, y 42.4, w 2.5, h 5.8 [#76B900]: green padlock at top-right of Sentry panel (outline padlock, stroke about 3 px, overlaps panel top-right corner)
- photo at x 40.2, y 32.5, w 19.6, h 63.9 [#1A1A1A]: Vera MGX compute node rack, vertical, dark chassis with glowing chips and silver base plate (unboxed, uncropped, sits on black; dominant black, silver and gold-tan with blue and magenta chip accents)
- shape at x 42.9, y 44.3, w 5.3, h 2.4 [#000000]: Vera CPU label box (black fill, thin white outline, sits over rack chassis)
- shape at x 51.1, y 70.2, w 5.7, h 2.4 [#000000]: BlueField-4 label box (black fill, thin white outline, sits over rack chassis)
- line at x 33.6, y 49.5, w 9, h 0.8 [#76B900]: horizontal double-headed arrow from OpenShell panel to rack (about 2 px stroke with arrowheads at both ends)
- line at x 43.8, y 52.8, w 0.8, h 19 [#76B900]: vertical double-headed arrow inside rack, upper chips to lower chips (about 2 px stroke with arrowheads at both ends)
- line at x 45.6, y 75.7, w 20.9, h 0.8 [#76B900]: horizontal arrow from rack to Sentry panel (about 2 px stroke, arrowhead pointing right at Sentry panel)
- text at x 59.9, y 64.8, w 4, h 11 [#757575]: Prompt / Tools / Data / Model / Policies (stacked centred grey labels between rack and Sentry panel)
- shape at x 68.2, y 51.3, w 23.1, h 5.3 [#262626]: Hardware-Based Enforcement tile (flat fill, slight corner rounding, no shadow)
- shape at x 68.2, y 57.9, w 23.1, h 5.3 [#262626]: Continuous Agent Monitoring tile (flat fill, slight corner rounding, no shadow)
- shape at x 68.2, y 64.5, w 23.1, h 5.3 [#262626]: Trusted Telemetry tile (flat fill, slight corner rounding, no shadow)
- shape at x 68.2, y 71.3, w 23.1, h 5.3 [#262626]: Millisecond-Scale Containment tile (flat fill, slight corner rounding, no shadow)
- shape at x 8.5, y 53.3, w 10.3, h 5.3 [#262626]: Gateway tile (flat fill, slight corner rounding, no shadow)
- shape at x 8.5, y 59.9, w 10.3, h 5.3 [#262626]: Supervisor tile (flat fill, slight corner rounding, no shadow)
- shape at x 8.5, y 66.5, w 10.3, h 5.3 [#262626]: Policy Prover tile (flat fill, slight corner rounding, no shadow)
- shape at x 20.6, y 54.8, w 10.3, h 13.3 [#262626]: Agent Harness tile (flat fill, slight corner rounding, no shadow)
- line at x 69.6, y 80, w 0.6, h 5.3 [#757575]: grey dashed double-headed arrow above Prompt tile (dashed about 1.5 px, arrowheads both ends)
- line at x 76.1, y 80, w 0.6, h 5.3 [#757575]: grey dashed double-headed arrow above Tools tile (dashed about 1.5 px, arrowheads both ends)
- line at x 82.7, y 80, w 0.6, h 5.3 [#757575]: grey dashed double-headed arrow above Data tile (dashed about 1.5 px, arrowheads both ends)
- line at x 89.3, y 80, w 0.6, h 5.3 [#757575]: grey dashed double-headed arrow above Model tile (dashed about 1.5 px, arrowheads both ends)
- shape at x 66.8, y 86.5, w 6.1, h 5.2 [#262626]: Prompt tile (flat fill, slight corner rounding)
- shape at x 73.3, y 86.5, w 6.2, h 5.2 [#262626]: Tools tile (flat fill, slight corner rounding)
- shape at x 79.9, y 86.5, w 6.2, h 5.2 [#262626]: Data tile (flat fill, slight corner rounding)
- shape at x 86.5, y 86.5, w 6.1, h 5.2 [#262626]: Model tile (flat fill, slight corner rounding)
- logo at x 92.9, y 94.2, w 4.7, h 1.6 [#76B900]: NVIDIA logo, green mark with white wordmark (bottom-right corner logo)
- title at x 16.9, y 10.5, w 66.2, h 4.9 [NVIDIA Sans Medium, regular weight, sentence-style caps; font size 4.94% of slide height; #FFFFFF; center aligned]: Announcing NVIDIA Open Agent Security Platform
- subtitle at x 31.8, y 16.1, w 36.5, h 2 [NVIDIA Sans Medium, regular weight; font size 1.98% of slide height; #FFFFFF; center aligned]: An Open Reference Design with NVIDIA OpenShell and NVIDIA Sentry
- body at x 9.1, y 19.6, w 81.9, h 2 [NVIDIA Sans Regular, regular weight; font size 1.98% of slide height; #A1A1A1; center aligned]: OpenShell is a policy-controlled runtime between an agent and the enterprise systems it can affect – data, files, credentials, tools, APIs, models, and networks.
- body at x 7.4, y 23.7, w 85.5, h 7.6 [NVIDIA Sans Regular, regular weight; font size 1.98% of slide height; #A1A1A1; center aligned]: OpenShell addresses the agent security problem: 1) Prevents uncontrolled access to enterprise data, 2) Limits credential exposure and misuse, 3) Controls agent tool / use and code execution, 4) Applies default-deny network egress where appropriate, 5) Helps contain prompt injection, malicious retrieved content, / and harmful tool calls, and 6) Creates an auditable runtime boundary for long-running autonomous workflows. (three centred lines with line spacing about 2.6 percent of slide height)
- title at x 12.5, y 35.5, w 15.3, h 2.2 [NVIDIA Sans Medium, regular weight; font size 2.22% of slide height; #FFFFFF; center aligned]: Secure Runtime Boundary
- subtitle at x 6.3, y 38.2, w 27.8, h 1.7 [NVIDIA Sans Regular, regular weight; font size 1.73% of slide height; #FFFFFF; center aligned]: Governs and Mediates All Actions, Enforces Policy at Runtime
- title at x 70.5, y 35.5, w 18.5, h 2.2 [NVIDIA Sans Medium, regular weight; font size 2.22% of slide height; #FFFFFF; center aligned]: In-Silicon Out-of-Band Monitor
- subtitle at x 66.8, y 38.2, w 26, h 1.7 [NVIDIA Sans Regular, regular weight; font size 1.73% of slide height; #FFFFFF; center aligned]: Agent Reasoning Inspection and Tamper-Proof Telemetry
- title at x 15.4, y 47.7, w 9.6, h 2 [NVIDIA Sans Bold, regular case; font size 1.98% of slide height; #76B900; center aligned]: NVIDIA OpenShell
- title at x 75.9, y 47.7, w 7.7, h 2 [NVIDIA Sans Bold, regular case; font size 1.98% of slide height; #76B900; center aligned]: NVIDIA Sentry
- text at x 43.4, y 44.7, w 4.3, h 1.7 [NVIDIA Sans Regular, regular weight; font size 1.73% of slide height; #FFFFFF; center aligned]: Vera CPU
- text at x 51.5, y 70.4, w 5.1, h 1.7 [NVIDIA Sans Regular, regular weight; font size 1.73% of slide height; #FFFFFF; center aligned]: BlueField-4
- text at x 72.9, y 53.1, w 13.8, h 1.7 [NVIDIA Sans Regular, regular weight; font size 1.73% of slide height; #FFFFFF; center aligned]: Hardware-Based Enforcement
- text at x 73, y 58.7, w 13.8, h 3.8 [NVIDIA Sans Regular, regular weight; font size 1.73% of slide height; #FFFFFF; center aligned]: Continuous Agent Monitoring / and Threat Detection
- text at x 74.6, y 65.3, w 10.3, h 3.8 [NVIDIA Sans Regular, regular weight; font size 1.73% of slide height; #FFFFFF; center aligned]: Trusted Telemetry / and Detection Pipeline
- text at x 72.7, y 72.1, w 14.3, h 3.8 [NVIDIA Sans Regular, regular weight; font size 1.73% of slide height; #FFFFFF; center aligned]: Millisecond-Scale Containment / and Quarantine
- text at x 11.7, y 55.1, w 4, h 1.7 [NVIDIA Sans Regular, regular weight; font size 1.73% of slide height; #FFFFFF; center aligned]: Gateway
- text at x 11.2, y 61.7, w 4.9, h 1.7 [NVIDIA Sans Regular, regular weight; font size 1.73% of slide height; #FFFFFF; center aligned]: Supervisor
- text at x 10.7, y 68.3, w 5.9, h 1.7 [NVIDIA Sans Regular, regular weight; font size 1.73% of slide height; #FFFFFF; center aligned]: Policy Prover
- text at x 22.4, y 60.7, w 6.8, h 1.7 [NVIDIA Sans Regular, regular weight; font size 1.73% of slide height; #FFFFFF; center aligned]: Agent Harness
- text at x 24.9, y 69.2, w 6.1, h 1.5 [NVIDIA Sans Bold, regular case; font size 1.48% of slide height; #FFFFFF; right aligned]: Agent Sandbox
- text at x 68.1, y 88.3, w 3.5, h 1.7 [NVIDIA Sans Regular, regular weight; font size 1.73% of slide height; #FFFFFF; center aligned]: Prompt
- text at x 75.2, y 88.3, w 2.4, h 1.7 [NVIDIA Sans Regular, regular weight; font size 1.73% of slide height; #FFFFFF; center aligned]: Tools
- text at x 81.9, y 88.3, w 2.2, h 1.7 [NVIDIA Sans Regular, regular weight; font size 1.73% of slide height; #FFFFFF; center aligned]: Data
- text at x 88.2, y 88.3, w 2.8, h 1.7 [NVIDIA Sans Regular, regular weight; font size 1.73% of slide height; #FFFFFF; center aligned]: Model
- caption at x 44.1, y 93.2, w 11.7, h 1.7 [NVIDIA Sans Medium, regular weight; font size 1.73% of slide height; #FFFFFF; center aligned]: Vera MGX Compute Node
- caption at x 45.1, y 95.2, w 9.9, h 1.5 [NVIDIA Sans Regular, regular weight; font size 1.48% of slide height; #A1A1A1; center aligned]: Secure Agent Processing

**Measured from the file** (exact, read from the PDF rather than estimated; x, y, w, h as % of the slide; font size as % of the slide height)

- fill at x 0.0, y 0.0, w 100.0, h 100.0: #000000 (covers the slide)
- fill at x 8.5, y 66.5, w 10.3, h 5.3: #262626
- fill at x 8.5, y 59.9, w 10.3, h 5.3: #262626
- fill at x 8.5, y 53.3, w 10.3, h 5.3: #262626
- fill at x 20.6, y 54.8, w 10.3, h 13.3: #262626
- fill at x 68.2, y 64.5, w 23.1, h 5.3: #262626
- fill at x 68.2, y 57.9, w 23.1, h 5.3: #262626
- fill at x 68.2, y 51.3, w 23.1, h 5.3: #262626
- fill at x 68.2, y 71.3, w 23.1, h 5.3: #262626
- image at x 40.2, y 32.5, w 19.6, h 63.9: 1352 by 2475 pixels
- image at x 44.9, y 75.1, w 22.3, h 2.5: 1335 by 85 pixels
- image at x 43.4, y 51.7, w 1.4, h 22.3: 83 by 754 pixels
- image at x 32.9, y 48.9, w 10.4, h 2.5: 624 by 85 pixels
- text at x 16.9, y 10.5, w 66.2, h 4.9 [NVIDIASans-Medium; font size 4.94; #FFFFFF]: Announcing NVIDIA Open Agent Security Platform
- text at x 31.8, y 16.1, w 36.5, h 2.0 [NVIDIASans-Medium; font size 1.98; #FFFFFF]: An Open Reference Design with NVIDIA OpenShell and NVIDIA Sentry
- text at x 9.1, y 19.6, w 81.9, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: OpenShell is a policy-controlled runtime between an agent and the enterprise systems it can affect – data, files, credentials, tools, APIs, models, and networks.
- text at x 7.4, y 23.7, w 85.5, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: OpenShell addresses the agent security problem: 1) Prevents uncontrolled access to enterprise data, 2) Limits credential exposure and misuse, 3) Controls agent tool
- text at x 12.3, y 26.3, w 75.8, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: use and code execution, 4) Applies default-deny network egress where appropriate, 5) Helps contain prompt injection, malicious retrieved content,
- text at x 21.5, y 28.9, w 57.0, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: and harmful tool calls, and 6) Creates an auditable runtime boundary for long-running autonomous workflows.
- text at x 12.5, y 35.5, w 15.3, h 2.2 [NVIDIASans-Medium; font size 2.22; #FFFFFF]: Secure Runtime Boundary
- text at x 70.5, y 35.5, w 18.5, h 2.2 [NVIDIASans-Medium; font size 2.22; #FFFFFF]: In-Silicon Out-of-Band Monitor
- text at x 6.3, y 38.2, w 27.8, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: Governs and Mediates All Actions, Enforces Policy at Runtime
- text at x 66.8, y 38.2, w 26.0, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: Agent Reasoning Inspection and Tamper-Proof Telemetry
- text at x 15.4, y 47.7, w 9.6, h 2.0 [NVIDIASans-Bold; font size 1.98; #76B900]: NVIDIA OpenShell
- text at x 75.9, y 47.7, w 7.7, h 2.0 [NVIDIASans-Bold; font size 1.98; #76B900]: NVIDIA Sentry
- text at x 43.4, y 44.7, w 4.3, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: Vera CPU
- text at x 72.9, y 53.1, w 13.8, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: Hardware-Based Enforcement
- text at x 11.7, y 55.1, w 4.0, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: Gateway
- text at x 73.0, y 58.7, w 13.8, h 3.8 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: Continuous Agent Monitoring and Threat Detection
- text at x 22.4, y 60.7, w 6.8, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: Agent Harness
- text at x 11.2, y 61.7, w 4.9, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: Supervisor
- text at x 59.9, y 64.8, w 3.5, h 3.8 [NVIDIASans-Regular; font size 1.73; #757575]: PromptTools
- text at x 74.6, y 65.3, w 10.3, h 3.8 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: Trusted Telemetry and Detection Pipeline
- text at x 10.7, y 68.3, w 5.9, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: Policy Prover
- text at x 60.6, y 68.9, w 2.2, h 1.7 [NVIDIASans-Regular; font size 1.73; #757575]: Data
- text at x 51.5, y 70.4, w 5.1, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: BlueField-4
- text at x 60.3, y 71.0, w 2.8, h 1.7 [NVIDIASans-Regular; font size 1.73; #757575]: Model
- text at x 59.9, y 73.1, w 3.4, h 1.7 [NVIDIASans-Regular; font size 1.73; #757575]: Policies
- text at x 72.7, y 72.1, w 14.3, h 3.8 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: Millisecond-Scale Containment and Quarantine
- text at x 24.9, y 69.2, w 6.1, h 1.5 [NVIDIASans-Bold; font size 1.48; #FFFFFF]: Agent Sandbox
- text at x 68.1, y 88.3, w 3.5, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: Prompt
- text at x 75.2, y 88.3, w 2.4, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: Tools
- text at x 81.9, y 88.3, w 2.2, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: Data
- text at x 88.2, y 88.3, w 2.8, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: Model
- text at x 44.1, y 93.2, w 11.7, h 1.7 [NVIDIASans-Medium; font size 1.73; #FFFFFF]: Vera MGX Compute Node
- text at x 45.1, y 95.2, w 9.9, h 1.5 [NVIDIASans-Regular; font size 1.48; #A1A1A1]: Secure Agent Processing

### Slide 14

**Text**

NVIDIA Vera Delivers / Maximum Single-Threaded / Performance at Scale / NVIDIA’s data center CPU journey began in 2021 with the / introduction of the Grace CPU. / Today, we are in full production of our next-generation Vera CPU. / As a stand-alone product, Vera expands NVIDIA’s SAM even further. / CPUs have historically traded speed against volume: how fast one / task finishes versus how many run at once. Pre-cloud CPUs chose / speed, while cloud CPUs chose volume (more but slower cores). / Agentic AI needs both, and Vera is the first data center CPU / purpose-built for these workloads. / High / NEW / Pre-Cloud CPU / Frequency-Optimized / Max Single-Threaded / CPU at Scale / Single-Threaded / Core Performance / Low / Cloud CPU / Scale Out / Low / High / Throughput

**Layout**

Two-column layout on a black slide: a left text column (title and four body paragraphs) starts at about 5% from the left edge and occupies roughly the top-left to middle-left area, and a right-hand 2x2 matrix chart spans about 58% to 93% of the slide width. The title sits top-left, left-aligned at about 22% of slide height, with the body text stacked beneath it; the matrix is vertically centred on the right with its axis labels outside its left and bottom edges. The NVIDIA logo sits bottom-right at about 93% x and 94% y, with no page number or footer text visible.

**Visuals**

Diagram: a 2x2 quadrant matrix drawn with thin light-grey (#757575) lines, outer frame about 58.3% to 92.6% of slide width and 18.0% to 75.1% of slide height, with a vertical divider at about 75.4% of width and a horizontal divider at about 46.3% of height. Y-axis labelled High (top) and Low (bottom) with the green rotated-free label Single-Threaded Core Performance placed between them on the left; X-axis labelled Low (left column) and High (right column) beneath the frame, with the green label Throughput centred below. Top-left quadrant: small grey filled dot (#757575, about 1% diameter) above the grey label Pre-Cloud CPU and italic grey Frequency-Optimized. Top-right quadrant: highlighted box with a bright green (#76B900) border about 0.3% wide and a dark green tint fill (#598B00 at 15% opacity over black, appearing about #0E1A00); small bright green filled dot (#76B900) above green label NEW, then white label Max Single-Threaded / CPU at Scale. Bottom-right quadrant: small grey filled dot (#757575) above grey label Cloud CPU and italic grey Scale Out. Bottom-left quadrant is empty. No photographs, icons, or charts beyond this matrix; the only logo is the NVIDIA wordmark with its eye symbol in green (#76B900) and white text at bottom right.

**Design**

Background: Solid black #000000 across the entire slide. Pure black background (#000000) with no gradient or texture. Title text is white (#FFFFFF) NVIDIASans Medium, a clean geometric humanist sans-serif, approximately 4.9% of slide height per line with tight leading (about 5.3% line pitch), left-aligned. Body text is light grey (#A1A1A1) NVIDIASans Regular sans-serif at about 2.0% of slide height with generous line spacing (about 2.6% pitch between lines within a paragraph, paragraphs separated by about 1.0% extra space). Accent colour is NVIDIA green #76B900 used for the green axis label, NEW tag, and highlight border and dot. Matrix labels are white (#FFFFFF) for High/Low axis ends and quadrant names in the highlighted box, grey (#757575) for the non-highlighted quadrant labels, and green (#76B900) for the axis titles. Lines are thin grey strokes (#757575, roughly 0.1% of slide height) forming a square-cornered grid with no radius and no shadows; the highlighted quadrant has a green (#76B900) outline about 0.3% of slide height wide with square corners.

**Elements and positions** (back to front; x, y, width, height as % of the slide; type, colour and alignment in brackets)

- shape at x 0, y 0, w 100, h 100: Full-slide background fill, solid black (fill #000000, no border, covers entire slide; back-most element)
- shape at x 75.5, y 18.3, w 16.9, h 27.8: Highlight panel behind the top-right quadrant (fill #598B00 at 15% opacity over black (appears about #0E1A00); overlaps the top-right grid cell; square corners; outlined by the green border below)
- line at x 58.3, y 18, w 34.3, h 57.1: Matrix outer frame and dividers, thin grey grid (stroke #757575, approx 0.1% of slide height; square corners; vertical divider at x 75.4% spanning full height, horizontal divider at y 46.3% spanning full width)
- shape at x 75.5, y 18.3, w 16.9, h 27.8: Green highlight border around top-right quadrant (stroke #76B900, approx 0.3% of slide height; no fill; square corners; overlaps the grey grid lines on its edges)
- shape at x 66.1, y 26.3, w 1, h 1: Grey dot marking Pre-Cloud CPU (filled circle #757575, about 1% diameter, centred above the Pre-Cloud CPU label)
- shape at x 82.9, y 26.3, w 1, h 1: Green dot marking Max Single-Threaded CPU at Scale (filled circle #76B900, about 1% diameter, centred above the highlighted label)
- shape at x 84.9, y 56.5, w 1, h 1: Grey dot marking Cloud CPU (filled circle #757575, about 1% diameter, centred above the Cloud CPU label)
- title at x 5.2, y 22.2, w 35.8, h 15.5 [NVIDIASans, Medium weight, sans-serif, normal case, no letter-spacing; font size 4.94% of slide height; #FFFFFF; left aligned]: NVIDIA Vera Delivers / Maximum Single-Threaded / Performance at Scale (three lines, line pitch about 5.3% of slide height)
- body at x 5.2, y 41, w 35, h 29.2 [NVIDIASans, Regular weight, sans-serif, normal case, no letter-spacing; font size 1.98% of slide height; #A1A1A1; left aligned]: NVIDIA’s data center CPU journey began in 2021 with the / introduction of the Grace CPU. / Today, we are in full production of our next-generation Vera CPU. / As a stand-alone product, Vera expands NVIDIA’s SAM even further. / CPUs have historically traded speed against volume: how fast one / task finishes versus how many run at once. Pre-cloud CPUs chose / speed, while cloud CPUs chose volume (more but slower cores). / Agentic AI needs both, and Vera is the first data center CPU / purpose-built for these workloads. (four paragraphs separated by about 1% of slide height; line pitch about 2.6%; ragged right)
- logo at x 92.9, y 94, w 4.6, h 1.8 [#76B900]: NVIDIA wordmark logo with eye symbol, bottom right (green eye symbol (#76B900) with white bold NVIDIA text (#FFFFFF) to its right; no background)
- text at x 46, y 44, w 10.6, h 4.9 [NVIDIASans, Regular weight, sans-serif, normal case; font size 2.22% of slide height; #76B900; center aligned]: Single-Threaded / Core Performance (y-axis title, two lines, rotation none, sits left of the grid)
- text at x 54.3, y 31.7, w 2.7, h 2.2 [NVIDIASans, Regular weight, sans-serif, normal case; font size 2.22% of slide height; #FFFFFF; right aligned]: High (y-axis label aligned with top row)
- text at x 54.7, y 59.9, w 2.4, h 2.2 [NVIDIASans, Regular weight, sans-serif, normal case; font size 2.22% of slide height; #FFFFFF; right aligned]: Low (y-axis label aligned with bottom row)
- text at x 82.7, y 22.4, w 2.6, h 2 [NVIDIASans, Medium weight, sans-serif, capitals, no letter-spacing; font size 1.98% of slide height; #76B900; center aligned]: NEW (tag above the highlighted dot)
- text at x 60.7, y 30.1, w 12.8, h 5 [NVIDIASans, Medium weight for line 1 and Medium italic for line 2, sans-serif, normal case; font size 2.22% of slide height; #757575; center aligned]: Pre-Cloud CPU / Frequency-Optimized (line 1 Pre-Cloud CPU at y 30.1%; line 2 Frequency-Optimized in italic at y 32.8%)
- text at x 77.8, y 30.1, w 12.8, h 4.9 [NVIDIASans, Medium weight, sans-serif, normal case; font size 2.22% of slide height; #FFFFFF; center aligned]: Max Single-Threaded / CPU at Scale (two lines centred in highlighted quadrant)
- text at x 82.8, y 60, w 6.3, h 4.9 [NVIDIASans, Medium weight for Cloud CPU and Medium italic for Scale Out, sans-serif, normal case; font size 2.22% of slide height; #757575; center aligned]: Cloud CPU / Scale Out (line 1 Cloud CPU at y 60.0%; line 2 Scale Out in italic at y 62.7%)
- text at x 65.7, y 77.6, w 2.4, h 2.2 [NVIDIASans, Regular weight, sans-serif, normal case; font size 2.22% of slide height; #FFFFFF; center aligned]: Low (x-axis label under left column)
- text at x 82.6, y 77.6, w 2.7, h 2.2 [NVIDIASans, Regular weight, sans-serif, normal case; font size 2.22% of slide height; #FFFFFF; center aligned]: High (x-axis label under right column)
- text at x 72.1, y 81.9, w 6.9, h 2.2 [NVIDIASans, Regular weight, sans-serif, normal case; font size 2.22% of slide height; #76B900; center aligned]: Throughput (x-axis title centred below the axis labels)

**Measured from the file** (exact, read from the PDF rather than estimated; x, y, w, h as % of the slide; font size as % of the slide height)

- fill at x 0.0, y 0.0, w 100.0, h 100.0: #000000 (covers the slide)
- fill at x 75.5, y 18.3, w 16.9, h 27.8: #598B00, opacity 0.15
- text at x 5.2, y 22.2, w 27.9, h 4.9 [NVIDIASans-Medium; font size 4.94; #FFFFFF]: NVIDIA Vera Delivers
- text at x 5.2, y 27.5, w 35.8, h 4.9 [NVIDIASans-Medium; font size 4.94; #FFFFFF]: Maximum Single-Threaded
- text at x 5.3, y 32.8, w 28.2, h 4.9 [NVIDIASans-Medium; font size 4.94; #FFFFFF]: Performance at Scale
- text at x 82.7, y 22.4, w 2.6, h 2.0 [NVIDIASans-Medium; font size 1.98; #76B900]: NEW
- text at x 62.5, y 30.1, w 8.8, h 2.2 [NVIDIASans-Medium; font size 2.22; #757575]: Pre-Cloud CPU
- text at x 77.8, y 30.1, w 12.8, h 2.2 [NVIDIASans-Medium; font size 2.22; #FFFFFF]: Max Single-Threaded
- text at x 54.3, y 31.7, w 2.7, h 2.2 [NVIDIASans-Regular; font size 2.22; #FFFFFF]: High
- text at x 60.7, y 32.8, w 12.4, h 2.2 [NVIDIASans-Medium; font size 2.22; #757575]: Frequency-Optimized
- text at x 80.2, y 32.8, w 7.6, h 2.2 [NVIDIASans-Medium; font size 2.22; #FFFFFF]: CPU at Scale
- text at x 46.5, y 44.0, w 9.8, h 2.2 [NVIDIASans-Regular; font size 2.22; #76B900]: Single-Threaded
- text at x 5.2, y 41.0, w 29.9, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: NVIDIA’s data center CPU journey began in 2021 with the
- text at x 46.0, y 46.7, w 10.6, h 2.2 [NVIDIASans-Regular; font size 2.22; #76B900]: Core Performance
- text at x 5.2, y 43.9, w 16.2, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: introduction of the Grace CPU.
- text at x 5.2, y 48.4, w 33.7, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: Today, we are in full production of our next-generation Vera CPU.
- text at x 5.3, y 51.4, w 35.0, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: As a stand-alone product, Vera expands NVIDIA’s SAM even further.
- text at x 54.7, y 59.9, w 2.4, h 2.2 [NVIDIASans-Regular; font size 2.22; #FFFFFF]: Low
- text at x 82.8, y 60.0, w 6.3, h 2.2 [NVIDIASans-Medium; font size 2.22; #757575]: Cloud CPU
- text at x 5.3, y 55.8, w 34.2, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: CPUs have historically traded speed against volume: how fast one
- text at x 83.2, y 62.7, w 5.5, h 2.2 [NVIDIASans-Medium; font size 2.22; #757575]: Scale Out
- text at x 5.3, y 58.8, w 34.3, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: task finishes versus how many run at once. Pre-cloud CPUs chose
- text at x 5.2, y 61.7, w 32.6, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: speed, while cloud CPUs chose volume (more but slower cores).
- text at x 5.2, y 66.2, w 31.2, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: Agentic AI needs both, and Vera is the first data center CPU
- text at x 5.2, y 69.1, w 18.2, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: purpose-built for these workloads.
- text at x 65.7, y 77.6, w 2.4, h 2.2 [NVIDIASans-Regular; font size 2.22; #FFFFFF]: Low
- text at x 82.6, y 77.6, w 2.7, h 2.2 [NVIDIASans-Regular; font size 2.22; #FFFFFF]: High
- text at x 72.1, y 81.9, w 6.9, h 2.2 [NVIDIASans-Regular; font size 2.22; #76B900]: Throughput

### Slide 15

**Review notes**

- What would change the score: I would rebuild slide 28 as a visual summary tied to the three growth pillars, and cut the intro paragraphs above the charts on slides 11 and 15 to one line each.

**Text**

NVIDIA Vera Delivers Over 2X More Performance
and Memory Bandwidth for a 100 MW CPU Data Center
Vera completes agentic tasks 2.2X faster on the SPEC benchmark and provides 2.5X more bandwidth than leading x86 data center CPUs. With shipments
already underway to lead partners including OCI, SpaceXAI and AWS, Vera is expected to be deployed by every major hyperscaler, neocloud, AI lab, and
system OEM. With CPU revenue expected to more than double in FY2028, NVIDIA is set to become one of the world’s leading CPU suppliers.
100 MW CPU Data Center
x86 CPU 128C
NVIDIA Vera 88C
2.5X
2.0X
1.5X
1.0X
0.5X
0.0X
2.4X
2.1X
1.2X
210
7,600
180
140
3,600
90
CPUs (k)
Throughput
(SPECrate2026_int_base)
Memory Bandwidth (PB/s)
NVIDIA

**Layout**

Single centred column on a full-black slide: a two-line centred title near the top, a three-line centred grey subtitle beneath it, and one bordered chart panel centred in the lower-middle area. The NVIDIA logo sits in the bottom-right corner. No footer text or slide number is visible. Margins are about 10 percent on the left and right sides.

**Visuals**

Grouped bar chart titled 100 MW CPU Data Center, set inside a thin-outlined rounded rectangle panel. Legend above the plot: grey square followed by x86 CPU 128C, then green square followed by NVIDIA Vera 88C. Y axis runs from 0.0X to 2.5X in steps of 0.5X, labelled with X-suffixed values and no axis title. X axis categories are CPUs (k), Throughput (SPECrate2026_int_base) and Memory Bandwidth (PB/s). Series values: CPUs (k) x86 140 at 1.0X and Vera 180 at 1.2X; Throughput x86 3,600 at 1.0X and Vera 7,600 at 2.1X; Memory Bandwidth x86 90 at 1.0X and Vera 210 at 2.4X. Raw values are printed inside the bars in black bold or medium type, and multiplier labels sit above each bar in white. Grey bars are #989898 and green bars are #76B900, with a thin grey baseline. No highlighted points beyond the green series.

**Design**

Background: Solid black #000000 across the whole slide. Pure black background (#000000) with white (#FFFFFF) titles and chart labels and mid-grey (#A1A1A1) subtitle body text. Accent green #76B900 for the Vera series and logo. Typeface is NVIDIASans, a geometric humanist sans-serif, Medium for title and bold-ish labels, Regular for body. Title cap height is roughly 4.9 percent of slide height (about 5 percent font size), subtitle about 2.0 percent. Chart panel has a 1 px grey (#A1A1A1-like) outline with rounded corners of about 1 percent of slide height, no fill and no shadow. Bars are flat with no gradients or shadows.

**Elements and positions** (back to front; x, y, width, height as % of the slide; type, colour and alignment in brackets)

- shape at x 0, y 0, w 100, h 100 [#000000]: full-slide background rectangle (solid fill, no outline, covers the slide)
- shape at x 18.4, y 31, w 63.2, h 54.8 [#000000]: chart panel frame, rounded rectangle with no fill (1 px outline in grey around #A1A1A1, corner radius about 1 percent of slide height, no shadow)
- title at x 13.6, y 5.2, w 73.1, h 10.2 [NVIDIASans Medium, regular capitalisation, no letter-spacing; font size 4.94% of slide height; #FFFFFF; center aligned]: NVIDIA Vera Delivers Over 2X More Performance / and Memory Bandwidth for a 100 MW CPU Data Center (two lines, line spacing about 5.3 percent of slide height)
- body at x 10.3, y 18.1, w 79.7, h 7.2 [NVIDIASans Regular, normal weight, no caps, no letter-spacing; font size 1.98% of slide height; #A1A1A1; center aligned]: Vera completes agentic tasks 2.2X faster on the SPEC benchmark and provides 2.5X more bandwidth than leading x86 data center CPUs. With shipments / already underway to lead partners including OCI, SpaceXAI and AWS, Vera is expected to be deployed by every major hyperscaler, neocloud, AI lab, and / system OEM. With CPU revenue expected to more than double in FY2028, NVIDIA is set to become one of the world’s leading CPU suppliers. (three lines, line spacing about 2.6 percent of slide height)
- text at x 43.2, y 34.3, w 13.7, h 2 [NVIDIASans Medium, normal weight, no caps; font size 1.98% of slide height; #FFFFFF; center aligned]: 100 MW CPU Data Center
- shape at x 42, y 38.3, w 0.6, h 0.9 [#989898]: legend swatch for x86 CPU 128C
- text at x 42.7, y 38.2, w 5.6, h 1.5 [NVIDIASans Regular, normal weight; font size 1.48% of slide height; #FFFFFF; left aligned]: x86 CPU 128C
- shape at x 51.2, y 38.3, w 0.6, h 0.9 [#76B900]: legend swatch for NVIDIA Vera 88C
- text at x 51.9, y 38.2, w 6.5, h 1.5 [NVIDIASans Regular, normal weight; font size 1.48% of slide height; #FFFFFF; left aligned]: NVIDIA Vera 88C
- text at x 26.5, y 42.6, w 2, h 1.7 [NVIDIASans Regular, normal weight; font size 1.73% of slide height; #FFFFFF; right aligned]: 2.5X
- text at x 26.5, y 49.2, w 2, h 1.7 [NVIDIASans Regular, normal weight; font size 1.73% of slide height; #FFFFFF; right aligned]: 2.0X
- text at x 26.5, y 55.9, w 2, h 1.7 [NVIDIASans Regular, normal weight; font size 1.73% of slide height; #FFFFFF; right aligned]: 1.5X
- text at x 26.5, y 62.5, w 2, h 1.7 [NVIDIASans Regular, normal weight; font size 1.73% of slide height; #FFFFFF; right aligned]: 1.0X
- text at x 26.5, y 69.1, w 2, h 1.7 [NVIDIASans Regular, normal weight; font size 1.73% of slide height; #FFFFFF; right aligned]: 0.5X
- text at x 26.5, y 75.8, w 2, h 1.7 [NVIDIASans Regular, normal weight; font size 1.73% of slide height; #FFFFFF; right aligned]: 0.0X
- shape at x 29.5, y 76.5, w 44.9, h 0.1 [#8C8C8C]: chart baseline line (thin horizontal axis line, about 1 px)
- shape at x 33.3, y 63.4, w 3.6, h 13.3 [#989898]: bar x86 CPU 128C, CPUs (k), value 140
- shape at x 36.9, y 60.2, w 3.6, h 16.4 [#76B900]: bar NVIDIA Vera 88C, CPUs (k), value 180
- shape at x 48.3, y 63.4, w 3.6, h 13.3 [#989898]: bar x86 CPU 128C, Throughput, value 3,600
- shape at x 51.9, y 48.3, w 3.6, h 28.3 [#76B900]: bar NVIDIA Vera 88C, Throughput, value 7,600
- shape at x 63.3, y 63.4, w 3.6, h 13.3 [#989898]: bar x86 CPU 128C, Memory Bandwidth, value 90
- shape at x 66.9, y 44.6, w 3.6, h 32.1 [#76B900]: bar NVIDIA Vera 88C, Memory Bandwidth, value 210
- text at x 37.6, y 57.8, w 2.3, h 2 [NVIDIASans Medium, normal weight; font size 1.98% of slide height; #FFFFFF; center aligned]: 1.2X
- text at x 52.5, y 45.9, w 2.3, h 2 [NVIDIASans Medium, normal weight; font size 1.98% of slide height; #FFFFFF; center aligned]: 2.1X
- text at x 67.5, y 42.1, w 2.3, h 2 [NVIDIASans Medium, normal weight; font size 1.98% of slide height; #FFFFFF; center aligned]: 2.4X
- text at x 34.2, y 69.8, w 1.7, h 1.7 [NVIDIASans Medium, normal weight; font size 1.73% of slide height; #000000; center aligned]: 140
- text at x 37.8, y 66.5, w 1.7, h 1.7 [NVIDIASans Bold, normal weight; font size 1.73% of slide height; #000000; center aligned]: 180
- text at x 48.8, y 69.8, w 2.6, h 1.7 [NVIDIASans Medium, normal weight; font size 1.73% of slide height; #000000; center aligned]: 3,600
- text at x 52.4, y 60.7, w 2.6, h 1.7 [NVIDIASans Bold, normal weight; font size 1.73% of slide height; #000000; center aligned]: 7,600
- text at x 64.5, y 69.8, w 1.2, h 1.7 [NVIDIASans Medium, normal weight; font size 1.73% of slide height; #000000; center aligned]: 90
- text at x 67.8, y 53.6, w 1.7, h 1.7 [NVIDIASans Bold, normal weight; font size 1.73% of slide height; #000000; center aligned]: 210
- text at x 35, y 78.1, w 3.8, h 1.7 [NVIDIASans Regular, normal weight; font size 1.73% of slide height; #FFFFFF; center aligned]: CPUs (k)
- text at x 46, y 78.1, w 15, h 3.9 [NVIDIASans Regular, normal weight; font size 1.73% of slide height; #FFFFFF; center aligned]: Throughput / (SPECrate2026_int_base) (two lines, second line starts at y 80.3)
- text at x 61, y 78.1, w 11.9, h 1.7 [NVIDIASans Regular, normal weight; font size 1.73% of slide height; #FFFFFF; center aligned]: Memory Bandwidth (PB/s)
- logo at x 92.9, y 94.2, w 4.6, h 1.6 [#76B900]: NVIDIA logo, green eye mark with NVIDIA wordmark in white (wordmark in white #FFFFFF, placed bottom right)

**Measured from the file** (exact, read from the PDF rather than estimated; x, y, w, h as % of the slide; font size as % of the slide height)

- fill at x 0.0, y 0.0, w 100.0, h 100.0: #000000 (covers the slide)
- fill at x 33.3, y 63.4, w 33.6, h 13.3: #989898
- fill at x 36.9, y 44.6, w 33.6, h 32.1: #76B900
- text at x 18.2, y 5.2, w 63.4, h 4.9 [NVIDIASans-Medium; font size 4.94; #FFFFFF]: NVIDIA Vera Delivers Over 2X More Performance
- text at x 13.6, y 10.5, w 73.1, h 4.9 [NVIDIASans-Medium; font size 4.94; #FFFFFF]: and Memory Bandwidth for a 100 MW CPU Data Center
- text at x 10.3, y 18.1, w 79.7, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: Vera completes agentic tasks 2.2X faster on the SPEC benchmark and provides 2.5X more bandwidth than leading x86 data center CPUs. With shipments
- text at x 11.3, y 20.7, w 77.6, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: already underway to lead partners including OCI, SpaceXAI and AWS, Vera is expected to be deployed by every major hyperscaler, neocloud, AI lab, and
- text at x 13.8, y 23.3, w 72.7, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: system OEM. With CPU revenue expected to more than double in FY2028, NVIDIA is set to become one of the world’s leading CPU suppliers.
- text at x 43.2, y 34.3, w 13.7, h 2.0 [NVIDIASans-Medium; font size 1.98; #FFFFFF]: 100 MW CPU Data Center
- text at x 67.5, y 42.1, w 2.3, h 2.0 [NVIDIASans-Medium; font size 1.98; #FFFFFF]: 2.4X
- text at x 52.5, y 45.9, w 2.3, h 2.0 [NVIDIASans-Medium; font size 1.98; #FFFFFF]: 2.1X
- text at x 26.5, y 42.6, w 2.0, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: 2.5X
- text at x 42.7, y 38.2, w 5.6, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: x86 CPU 128C
- text at x 51.9, y 38.2, w 6.5, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: NVIDIA Vera 88C
- text at x 26.5, y 49.2, w 2.0, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: 2.0X
- text at x 37.6, y 57.8, w 2.3, h 2.0 [NVIDIASans-Medium; font size 1.98; #FFFFFF]: 1.2X
- text at x 67.8, y 53.6, w 1.7, h 1.7 [NVIDIASans-Bold; font size 1.73; #000000]: 210
- text at x 26.5, y 55.9, w 2.0, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: 1.5X
- text at x 52.4, y 60.7, w 2.6, h 1.7 [NVIDIASans-Bold; font size 1.73; #000000]: 7,600
- text at x 26.5, y 62.5, w 2.0, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: 1.0X
- text at x 37.8, y 66.5, w 1.7, h 1.7 [NVIDIASans-Bold; font size 1.73; #000000]: 180
- text at x 26.5, y 69.1, w 2.0, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: 0.5X
- text at x 34.2, y 69.8, w 1.7, h 1.7 [NVIDIASans-Medium; font size 1.73; #000000]: 140
- text at x 48.8, y 69.8, w 2.6, h 1.7 [NVIDIASans-Medium; font size 1.73; #000000]: 3,600
- text at x 64.5, y 69.8, w 1.2, h 1.7 [NVIDIASans-Medium; font size 1.73; #000000]: 90
- text at x 26.5, y 75.8, w 2.0, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: 0.0X
- text at x 35.0, y 78.1, w 3.8, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: CPUs (k)
- text at x 49.1, y 78.1, w 5.4, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: Throughput
- text at x 61.0, y 78.1, w 11.9, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: Memory Bandwidth (PB/s)
- text at x 46.0, y 80.3, w 11.6, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: (SPECrate2026_int_base)

### Slide 16

**Text**

Groq 3 LPX Co-Designed with Vera Rubin / Delivers Leadership Low-Latency Inference Performance / NVIDIA is in full production of Groq 3 LPX, the company’s first rack-scale LPU system. Together with the Vera Rubin NVL72 system, LPX expands the frontier / addressable by NVIDIA’s full-stack computing platform – specifically, the low-latency inference segment of the market. Groq 3 LPX is already setting records, / demonstrating nearly 4x the number of tokens per second against the next best alternative on Artificial Analysis’ Gemma 4 31B benchmark. / Throughput (TPS/MW) / Vera Rubin NVL72 / Blackwell / NVL72 / Hopper / NVL8 / LPX / Lowest Latency / Inference / User Interactivity (TPS / User) / Output Speed: Gemma 4 31B (Reasoning) / Output speed: o200k_base tokens per second · 100,000 input tokens / 3,431 / 4X Faster / 870 / 188 / 151 / 101 / 48 / 35 / 34 / 31 / 29 / NVIDIA Groq 3 LPX / Private, Pre-Release / Endpoint / Live Serverless Data / Public API Endpoints / Source: Artificial Analysis / NVIDIA logo

**Layout**

Two-column layout on a black field: a centred two-line title sits at the top, a centred three-line grey subtitle paragraph sits beneath it, and two equal-height outlined panels fill the lower two-thirds, the left panel (line chart) about 35% of the width and the right panel (bar chart) about 54% of the width, with a gutter of about 2% between them. Panel titles are centred at the top of each panel; the source line sits bottom-right inside the right panel; the NVIDIA logo sits bottom-right of the slide. There is no slide number or footer text.

**Visuals**

Two charts. (1) Left: a line-style throughput diagram titled 'Throughput (TPS/MW)' with x-axis 'User Interactivity (TPS / User)' and y-axis unlabelled, drawn as arrowed grey axes meeting at the bottom-left corner. Three curves: a green curve labelled 'Vera Rubin NVL72' starts high at the left axis, stays flat, then falls steeply to the baseline near x 23.5%; a white curve labelled 'Blackwell NVL72' starts at mid-height and falls to the baseline near x 13.4%; a short grey curve labelled 'Hopper NVL8' sits at the origin. A flat green segment labelled 'LPX' / 'Lowest Latency Inference' begins at the foot of the Vera Rubin drop and runs right and slightly down to the x-axis end. Two dashed green arrows: one diagonal from the Blackwell curve up toward the Vera Rubin curve, one horizontal from the LPX start pointing right. No values are printed on the axes. (2) Right: vertical bar chart titled 'Output Speed: Gemma 4 31B (Reasoning)' with subtitle 'Output speed: o200k_base tokens per second · 100,000 input tokens'. Y-axis is unlabelled with no printed scale; x-axis has no tick labels. Ten bars with printed values: NVIDIA Groq 3 LPX 3,431 (green, the highlighted bar, the tallest); 870; 188; 151; 101; 48; 35; 34; 31; 29 (grey). Values for the grey bars after 870 are printed; the bar heights for those are about proportional to value on a linear scale with 3,431 as full height. A green horizontal reference line runs from the top of the 3,431 bar to the right at the level of its top. A grey dashed double-headed vertical arrow runs from that reference line down to the top of the 870 bar, labelled '4X Faster' in green. Bar group labels: 'NVIDIA Groq 3 LPX' under the green bar, 'Private, Pre-Release / Endpoint' bold under it; 'Live Serverless Data / Public API Endpoints' under the grey bars. No legend.

**Design**

Background: Solid black #000000 covering the whole slide. Pure black background (#000000) with white (#FFFFFF) titles and chart labels, grey (#A1A1A1) subtitle body text, mid-grey (#757575) source line, light grey (#D9D9D9) secondary label, NVIDIA green (#76B900) for accents, the highlighted bar, the 4X Faster label, the Vera Rubin and LPX labels and curves. Typeface is NVIDIA Sans (a humanist sans-serif), Medium weight for titles and labels, Regular weight for the subtitle and source line. Title cap height about 4.9% of slide height (font size 4.94% of height); subtitle about 2.0% of height; chart labels about 1.2 to 1.5%. Panel outlines are thin grey strokes (about #6B6B6B, roughly 1px) with small corner radii (about 1% of slide height). Grey bars are #757575, flat with no shadow or gradient; axes are thin grey (#757575) lines with small arrowheads. No shadows or gradients.

**Elements and positions** (back to front; x, y, width, height as % of the slide; type, colour and alignment in brackets)

- shape at x 0, y 0, w 100, h 100 [#000000]: Full-slide background rectangle, black fill
- shape at x 4.95, y 31, w 34.7, h 54.8 [#000000]: Left chart panel outline, rounded rectangle, no fill (1px stroke about #6B6B6B, corner radius about 1% of slide height, no shadow)
- shape at x 41.15, y 31, w 53.85, h 54.8 [#000000]: Right chart panel outline, rounded rectangle, no fill (1px stroke about #6B6B6B, corner radius about 1% of slide height, no shadow)
- shape at x 45.4, y 42.5, w 3.4, h 34.1 [#76B900]: Bar 1: NVIDIA Groq 3 LPX bar, value 3,431, rectangle (Bottom edge sits on the x-axis baseline at y 76.5; no outline)
- shape at x 53.4, y 67.9, w 3.35, h 8.6 [#757575]: Bar 2: grey bar, value 870 (Bottom on the baseline; no outline)
- shape at x 57.9, y 74.8, w 3.35, h 1.7 [#757575]: Bar 3: grey bar, value 188 (Bottom on the baseline; no outline)
- shape at x 62.4, y 75.1, w 3.35, h 1.4 [#757575]: Bar 4: grey bar, value 151 (Bottom on the baseline; no outline)
- shape at x 66.9, y 75.6, w 3.35, h 0.9 [#757575]: Bar 5: grey bar, value 101 (Bottom on the baseline; no outline)
- shape at x 71.4, y 76.1, w 3.35, h 0.4 [#757575]: Bar 6: grey bar, value 48 (Bottom on the baseline; no outline)
- shape at x 75.9, y 76.3, w 3.35, h 0.2 [#757575]: Bar 7: grey bar, value 35 (Bottom on the baseline; no outline)
- shape at x 80.4, y 76.3, w 3.35, h 0.2 [#757575]: Bar 8: grey bar, value 34 (Bottom on the baseline; no outline)
- shape at x 84.9, y 76.3, w 3.35, h 0.2 [#757575]: Bar 9: grey bar, value 31 (Bottom on the baseline; no outline)
- shape at x 89.4, y 76.3, w 3.35, h 0.2 [#757575]: Bar 10: grey bar, value 29 (Bottom on the baseline; no outline)
- line at x 43.55, y 76.5, w 50.1, h 0.1 [#757575]: Bar chart x-axis baseline (Thin grey line, 1px)
- line at x 48.95, y 42.5, w 6.3, h 0.1 [#76B900]: Green reference line from top of 3,431 bar to the right (1px green line, horizontal)
- line at x 55, y 43.8, w 0.1, h 21.1 [#757575]: Grey dashed double-headed vertical arrow, 4X Faster marker (Dashed grey stroke about 1px with small filled arrowheads at both ends; interrupted around the 4X Faster label)
- line at x 8.45, y 40.9, w 0.1, h 35.6 [#757575]: Left chart y-axis, arrowed (Thin grey line with arrowhead at top)
- line at x 8.45, y 76.5, w 28.6, h 0.1 [#757575]: Left chart x-axis, arrowed (Thin grey line with arrowhead at right end)
- line at x 8.45, y 46.7, w 15.05, h 29.8 [#76B900]: Vera Rubin NVL72 curve (About 2px green stroke, curve starts high at left axis, flat then steep drop to baseline at x about 23.5)
- line at x 8.45, y 66.2, w 4.95, h 10.3 [#FFFFFF]: Blackwell NVL72 curve (About 2px white stroke, drops to baseline at x about 13.4)
- line at x 8.45, y 74.5, w 1.05, h 2 [#757575]: Hopper NVL8 curve (About 2px grey stroke, short curve at origin)
- line at x 22.5, y 66, w 13.4, h 10.5 [#76B900]: LPX segment, green curve (About 2px green stroke, begins at the Vera Rubin drop and runs right and gently down to the x-axis end)
- line at x 12.1, y 56.9, w 5.5, h 11.1 [#76B900]: Dashed green diagonal arrow (Dashed about 1.5px stroke, arrowhead at upper right pointing toward Vera Rubin curve)
- line at x 23.5, y 74.3, w 5.4, h 0.1 [#76B900]: Dashed green horizontal arrow (Dashed about 1.5px stroke, arrowhead pointing right)
- text at x 23.1, y 5.2, w 54.4, h 4.9 [NVIDIASans, Medium, sans-serif, normal case, no letter-spacing; font size 4.94% of slide height; #FFFFFF; center aligned]: Groq 3 LPX Co-Designed with Vera Rubin
- text at x 13, y 10.5, w 74.2, h 4.9 [NVIDIASans, Medium, sans-serif, normal case, no letter-spacing; font size 4.94% of slide height; #FFFFFF; center aligned]: Delivers Leadership Low-Latency Inference Performance
- body at x 9.5, y 18.1, w 81.2, h 7.2 [NVIDIASans, Regular, sans-serif, normal case; font size 1.98% of slide height; #A1A1A1; center aligned]: NVIDIA is in full production of Groq 3 LPX, the company’s first rack-scale LPU system. Together with the Vera Rubin NVL72 system, LPX expands the frontier / addressable by NVIDIA’s full-stack computing platform – specifically, the low-latency inference segment of the market. Groq 3 LPX is already setting records, / demonstrating nearly 4x the number of tokens per second against the next best alternative on Artificial Analysis’ Gemma 4 31B benchmark. (Three lines, line pitch about 2.6% of slide height; line 1 at x 9.5 w 81.2, line 2 at x 9.5 w 81.4, line 3 at x 13.7 w 72.8)
- text at x 16.4, y 34.3, w 11.9, h 2 [NVIDIASans, Medium, sans-serif, normal case; font size 1.98% of slide height; #FFFFFF; center aligned]: Throughput (TPS/MW)
- text at x 58, y 34.3, w 21.7, h 2 [NVIDIASans, Medium, sans-serif, normal case; font size 1.98% of slide height; #FFFFFF; center aligned]: Output Speed: Gemma 4 31B (Reasoning)
- text at x 55.4, y 37, w 26.9, h 1.5 [NVIDIASans, Regular, sans-serif, normal case; font size 1.48% of slide height; #FFFFFF; center aligned]: Output speed: o200k_base tokens per second · 100,000 input tokens
- text at x 9.1, y 43.6, w 8.2, h 1.7 [NVIDIASans, Medium, sans-serif, normal case; font size 1.73% of slide height; #76B900; left aligned]: Vera Rubin NVL72
- text at x 45.9, y 40.1, w 2.2, h 1.5 [NVIDIASans, Medium, sans-serif, normal case; font size 1.48% of slide height; #FFFFFF; center aligned]: 3,431
- text at x 52.5, y 53.4, w 5, h 2 [NVIDIASans, Medium, sans-serif, normal case; font size 1.98% of slide height; #76B900; left aligned]: 4X Faster
- text at x 29.2, y 65.5, w 1.8, h 1.7 [NVIDIASans, Medium, sans-serif, normal case; font size 1.73% of slide height; #76B900; left aligned]: LPX
- text at x 54.3, y 65.9, w 1.5, h 1.5 [NVIDIASans, Medium, sans-serif, normal case; font size 1.48% of slide height; #FFFFFF; center aligned]: 870
- text at x 29.2, y 68.3, w 6.3, h 3.1 [NVIDIASans, Regular, sans-serif, normal case; font size 1.48% of slide height; #FFFFFF; left aligned]: Lowest Latency / Inference (Two lines, left aligned)
- text at x 58.8, y 72.6, w 1.5, h 1.5 [NVIDIASans, Medium, sans-serif, normal case; font size 1.48% of slide height; #FFFFFF; center aligned]: 188
- text at x 63.3, y 73, w 1.5, h 1.5 [NVIDIASans, Medium, sans-serif, normal case; font size 1.48% of slide height; #FFFFFF; center aligned]: 151
- text at x 67.8, y 73.5, w 1.5, h 1.5 [NVIDIASans, Medium, sans-serif, normal case; font size 1.48% of slide height; #FFFFFF; center aligned]: 101
- text at x 72.6, y 74, w 1, h 1.5 [NVIDIASans, Medium, sans-serif, normal case; font size 1.48% of slide height; #FFFFFF; center aligned]: 48
- text at x 77.1, y 74.1, w 1, h 1.5 [NVIDIASans, Medium, sans-serif, normal case; font size 1.48% of slide height; #FFFFFF; center aligned]: 35
- text at x 81.6, y 74.2, w 1, h 1.5 [NVIDIASans, Medium, sans-serif, normal case; font size 1.48% of slide height; #FFFFFF; center aligned]: 34
- text at x 86.1, y 74.2, w 1, h 1.5 [NVIDIASans, Medium, sans-serif, normal case; font size 1.48% of slide height; #FFFFFF; center aligned]: 31
- text at x 90.6, y 74.2, w 1, h 1.5 [NVIDIASans, Medium, sans-serif, normal case; font size 1.48% of slide height; #FFFFFF; center aligned]: 29
- text at x 9.1, y 63, w 3.3, h 2.6 [NVIDIASans, Medium, sans-serif, normal case; font size 1.24% of slide height; #D9D9D9; left aligned]: Blackwell NVL72 / (two-line label: Blackwell / NVL72)
- text at x 16.6, y 79, w 11.5, h 1.5 [NVIDIASans, Regular, sans-serif, normal case; font size 1.48% of slide height; #FFFFFF; center aligned]: User Interactivity (TPS / User)
- text at x 84.2, y 82.9, w 9.8, h 1.5 [NVIDIASans, Regular, sans-serif, normal case; font size 1.48% of slide height; #757575; right aligned]: Source: Artificial Analysis
- text at x 9.1, y 71.5, w 2.6, h 2.6 [NVIDIASans, Medium, sans-serif, normal case; font size 1.24% of slide height; #757575; left aligned]: Hopper / NVL8
- text at x 44, y 78, w 6.1, h 1.2 [NVIDIASans, Regular, sans-serif, normal case; font size 1.24% of slide height; #FFFFFF; center aligned]: NVIDIA Groq 3 LPX
- text at x 43.7, y 80.4, w 6.6, h 1.2 [NVIDIASans, Medium, sans-serif, normal case; font size 1.24% of slide height; #FFFFFF; center aligned]: Private, Pre-Release
- text at x 45.5, y 81.9, w 3, h 1.2 [NVIDIASans, Medium, sans-serif, normal case; font size 1.24% of slide height; #FFFFFF; center aligned]: Endpoint
- text at x 69.8, y 80.4, w 6.6, h 1.2 [NVIDIASans, Medium, sans-serif, normal case; font size 1.24% of slide height; #FFFFFF; left aligned]: Live Serverless Data
- text at x 69.7, y 81.9, w 6.7, h 1.2 [NVIDIASans, Regular, sans-serif, normal case; font size 1.24% of slide height; #FFFFFF; left aligned]: Public API Endpoints
- logo at x 92.8, y 94, w 4.6, h 1.8 [#76B900]: NVIDIA logo: green eye mark with NVIDIA wordmark in white, bottom right corner (Wordmark in white (#FFFFFF), mark in green (#76B900))

**Measured from the file** (exact, read from the PDF rather than estimated; x, y, w, h as % of the slide; font size as % of the slide height)

- fill at x 0.0, y 0.0, w 100.0, h 100.0: #000000 (covers the slide)
- fill at x 45.4, y 42.5, w 3.4, h 34.1: #76B900
- text at x 23.1, y 5.2, w 54.4, h 4.9 [NVIDIASans-Medium; font size 4.94; #FFFFFF]: Groq 3 LPX Co-Designed with Vera Rubin
- text at x 13.0, y 10.5, w 74.2, h 4.9 [NVIDIASans-Medium; font size 4.94; #FFFFFF]: Delivers Leadership Low-Latency Inference Performance
- text at x 9.5, y 18.1, w 81.2, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: NVIDIA is in full production of Groq 3 LPX, the company’s first rack-scale LPU system. Together with the Vera Rubin NVL72 system, LPX expands the frontier
- text at x 9.5, y 20.7, w 81.4, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: addressable by NVIDIA’s full-stack computing platform – specifically, the low-latency inference segment of the market. Groq 3 LPX is already setting records,
- text at x 13.7, y 23.3, w 72.8, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: demonstrating nearly 4x the number of tokens per second against the next best alternative on Artificial Analysis’ Gemma 4 31B benchmark.
- text at x 16.4, y 34.3, w 11.9, h 2.0 [NVIDIASans-Medium; font size 1.98; #FFFFFF]: Throughput (TPS/MW)
- text at x 58.0, y 34.3, w 21.7, h 2.0 [NVIDIASans-Medium; font size 1.98; #FFFFFF]: Output Speed: Gemma 4 31B (Reasoning)
- text at x 55.4, y 37.0, w 26.9, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Output speed: o200k_base tokens per second · 100,000 input tokens
- text at x 9.1, y 43.6, w 8.2, h 1.7 [NVIDIASans-Medium; font size 1.73; #76B900]: Vera Rubin NVL72
- text at x 45.9, y 40.1, w 2.2, h 1.5 [NVIDIASans-Medium; font size 1.48; #FFFFFF]: 3,431
- text at x 52.5, y 53.4, w 5.0, h 2.0 [NVIDIASans-Medium; font size 1.98; #76B900]: 4X Faster
- text at x 29.2, y 65.5, w 1.8, h 1.7 [NVIDIASans-Medium; font size 1.73; #76B900]: LPX
- text at x 54.3, y 65.9, w 1.5, h 1.5 [NVIDIASans-Medium; font size 1.48; #FFFFFF]: 870
- text at x 29.2, y 68.3, w 6.3, h 3.1 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Lowest Latency Inference
- text at x 58.8, y 72.6, w 1.5, h 1.5 [NVIDIASans-Medium; font size 1.48; #FFFFFF]: 188
- text at x 63.3, y 73.0, w 1.5, h 1.5 [NVIDIASans-Medium; font size 1.48; #FFFFFF]: 151
- text at x 67.8, y 73.5, w 1.5, h 1.5 [NVIDIASans-Medium; font size 1.48; #FFFFFF]: 101
- text at x 72.6, y 74.0, w 1.0, h 1.5 [NVIDIASans-Medium; font size 1.48; #FFFFFF]: 48
- text at x 77.1, y 74.1, w 1.0, h 1.5 [NVIDIASans-Medium; font size 1.48; #FFFFFF]: 35
- text at x 81.6, y 74.2, w 1.0, h 1.5 [NVIDIASans-Medium; font size 1.48; #FFFFFF]: 34
- text at x 86.1, y 74.2, w 1.0, h 1.5 [NVIDIASans-Medium; font size 1.48; #FFFFFF]: 31
- text at x 90.6, y 74.2, w 1.0, h 1.5 [NVIDIASans-Medium; font size 1.48; #FFFFFF]: 29
- text at x 9.1, y 63.0, w 3.3, h 2.6 [NVIDIASans-Medium; font size 1.24; #D9D9D9]: Blackwell NVL72
- text at x 16.6, y 79.0, w 11.5, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: User Interactivity (TPS / User)
- text at x 84.2, y 82.9, w 9.8, h 1.5 [NVIDIASans-Regular; font size 1.48; #757575]: Source: Artificial Analysis
- text at x 9.1, y 71.5, w 2.6, h 2.6 [NVIDIASans-Medium; font size 1.24; #757575]: Hopper NVL8
- text at x 44.0, y 78.0, w 6.1, h 1.2 [NVIDIASans-Regular; font size 1.24; #FFFFFF]: NVIDIA Groq 3 LPX
- text at x 43.7, y 80.4, w 6.6, h 1.2 [NVIDIASans-Medium; font size 1.24; #FFFFFF]: Private, Pre-Release
- text at x 69.8, y 80.4, w 6.6, h 1.2 [NVIDIASans-Medium; font size 1.24; #FFFFFF]: Live Serverless Data
- text at x 45.5, y 81.9, w 3.0, h 1.2 [NVIDIASans-Medium; font size 1.24; #FFFFFF]: Endpoint
- text at x 69.7, y 81.9, w 6.7, h 1.2 [NVIDIASans-Regular; font size 1.24; #FFFFFF]: Public API Endpoints

### Slide 17

**Review notes**

- Worth studying: An indexed growth chart against the total market is labelled with a CAGR comparison, turning a single claim into a visible contrast.
- Review: The deck has a consistent black and NVIDIA green identity, action-style titles, and charts that carry real arguments, notably the paired curves on slide 10, the per-GW bars on slide 11 and the indexed networking chart on slide 17.
- Worth keeping: Keep the paired comparison charts on slide 10, the green-against-gray emphasis with annotated multiples, and the single-chart-with-numbered-list layout on slide 17.

**Text**

NVIDIA is the Largest and
Fastest Growing AI Networking
Company in the World
NVIDIA is the only company that offers five networking technologies:
1. Scale-Up (NVLink)
2. Scale-Out (InfiniBand, Spectrum-X Ethernet)
3. Scale-Across (Spectrum-XGS Ethernet)
4. Scale-In (BlueField-4 DPU and DOCA over Spectrum-X Ethernet)
5. Context Scale (Vera BlueField-4 Storage Processor)
Since the acquisition of Mellanox in 2020, NVIDIA's Networking revenue
has grown at a CAGR of 70%, materially outperforming industry peers.
Today, NVIDIA is the largest and fastest growing AI Networking company
in the world.
Scale-Up
NVLink
Scale-Out
Spectrum-X Ethernet
Scale-Across
Spectrum-XGS Ethernet
Scale-In
BlueField-4 DPU
Context Scale
Vera BlueField-4 Storage Processor
Networking Revenue (Indexed CY2020 = 100)
NVIDIA
Total Market
CAGR (CY2020-2025)
NVIDIA – 70%
Market – 28%
1,600
1,400
1,200
1,000
800
600
400
200
0
1435
594
392
169
100
110
139
180
218
347
100
109
2020
2021
2022
2023
2024
2025

**Layout**

Two-column layout on a black field. The left column (text block, x about 5% to 47%) is vertically centred, with the title at the top of that block and the body text below it, left-aligned to a shared margin at about 5.2% of slide width. The right column (x about 50.7% to 95.1%) holds a row of five product tiles across the top (y about 14.9% to 29.6%) above a single bordered chart panel (y about 31% to 85.8%). The NVIDIA logo sits bottom right at about x 92.9%, y 94.2%. No slide number or footer text is visible.

**Visuals**

Five product photographs in a row, each in a thin-bordered rounded rectangle, with a green label above and a small white caption below. (1) Scale-Up, labelled NVLink: a dark grey vertical NVLink switch/compute board with a heatsink and connectors, shown upright, centred on black, small framing with black surround. (2) Scale-Out, labelled Spectrum-X Ethernet: a dark metal networking switch chassis in a three-quarter angled view with a visible fan grille and a round top vent, lit with warm light, filling the tile and cropped at the edges. (3) Scale-Across, labelled Spectrum-XGS Ethernet: a diagram-style illustration on black of four small grey switch boxes arranged around a central box, connected by thin white lines, with small labels reading Scale-Up, Scale-Across and Scale-Out in tiny text. (4) Scale-In, labelled BlueField-4 DPU: a dark green-black circuit board with a central square chip, heatsinks and connectors, shown flat, black background. (5) Context Scale, labelled Vera BlueField-4 Storage Processor: a dark grey rack server chassis viewed from the front-top angle with an open lid, showing a green-lit board inside and rows of drive bays. Below the tiles, one bordered chart: a line chart titled Networking Revenue (Indexed CY2020 = 100). Legend above plot, centred: green line NVIDIA, grey line Total Market. Y axis: 0 to 1,600 in steps of 200, unitless index, no gridlines, tick labels at left. X axis: 2020, 2021, 2022, 2023, 2024, 2025 along a thin grey baseline at 0. Series NVIDIA (green #76B900, thick line, labelled values): 2020 100, 2021 110, 2022 169, 2023 392, 2024 594, 2025 1435. Series Total Market (grey #A1A1A1, labelled values): 2020 100, 2021 109, 2022 139, 2023 180, 2024 218, 2025 347. Labels are printed at each point; the 2025 NVIDIA point is highlighted in green with its label at the top. Inside the plot, at left-centre, a text callout reads CAGR (CY2020-2025) with NVIDIA – 70% and Market – 28% in white. Chart has no visible axis titles. Also the NVIDIA logo: green eye-shaped mark with the wordmark NVIDIA in white, bottom right.

**Design**

Background: Solid black #000000 covering the whole slide. Black background #000000. Titles white #FFFFFF in NVIDIASans Medium, a clean geometric humanist sans-serif, sentence-style capitalisation as printed. Title size about 4.9% of slide height with tight line spacing (about 1.1). Body text grey #A1A1A1 in NVIDIASans Regular, about 2.0% of slide height, with generous line spacing. Accent colour NVIDIA green #76B900 for list numbers, tile labels, chart NVIDIA series and its data labels. Tile captions are white #FFFFFF at about 1.0% of slide height. Chart panel and tile frames are outlined with thin grey strokes (about #5A5A5A, roughly 1 px) with small rounded corners (radius about 1% of slide width), no fill, no shadows, no gradients. Numbered list uses green numerals with a hanging indent and no bullets. Chart text in white or grey, chart data labels in green or grey, Medium weight for the callout and labels.

**Elements and positions** (back to front; x, y, width, height as % of the slide; type, colour and alignment in brackets)

- shape at x 0, y 0, w 100, h 100 [#000000]: Full-slide black background fill (Covers the slide, no border.)
- title at x 5.2, y 22.2, w 41.7, h 15.5 [NVIDIASans Medium, regular weight, sentence case; font size 4.94% of slide height; #FFFFFF; left aligned]: NVIDIA is the Largest and / Fastest Growing AI Networking / Company in the World (Three lines, line spacing about 1.1 of font size.)
- body at x 5.2, y 41, w 35.8, h 2 [NVIDIASans Regular, regular weight; font size 1.98% of slide height; #A1A1A1; left aligned]: NVIDIA is the only company that offers five networking technologies:
- bullet list at x 7.7, y 45.4, w 33.1, h 17.4 [NVIDIASans Regular, regular weight; font size 1.98% of slide height; #A1A1A1; left aligned]: 1. Scale-Up (NVLink) / 2. Scale-Out (InfiniBand, Spectrum-X Ethernet) / 3. Scale-Across (Spectrum-XGS Ethernet) / 4. Scale-In (BlueField-4 DPU and DOCA over Spectrum-X Ethernet) / 5. Context Scale (Vera BlueField-4 Storage Processor) (Numbered list, numerals in green #76B900 Regular, text grey #A1A1A1, item spacing about 4.1% of slide height, text indented at x 9.1%.)
- body at x 5.3, y 64.3, w 37.2, h 4.4 [NVIDIASans Regular, regular weight; font size 1.98% of slide height; #A1A1A1; left aligned]: Since the acquisition of Mellanox in 2020, NVIDIA's Networking revenue / has grown at a CAGR of 70%, materially outperforming industry peers.
- body at x 5.3, y 70.5, w 37.8, h 4.4 [NVIDIASans Regular, regular weight; font size 1.98% of slide height; #A1A1A1; left aligned]: Today, NVIDIA is the largest and fastest growing AI Networking company / in the world.
- photo at x 59.7, y 18.9, w 8.2, h 10.6 [#2A2A2A]: Scale-Up tile photograph: dark grey NVLink board with heatsink and connectors, upright, on black (Boxed in a rounded-corner tile with thin grey stroke, about 1% corner radius, black surround, image fully inside frame.)
- text at x 52.8, y 14.9, w 4.2, h 1.7 [NVIDIASans Medium, regular weight; font size 1.73% of slide height; #76B900; center aligned]: Scale-Up
- text at x 54, y 17, w 1.8, h 1 [NVIDIASans Regular, regular weight; font size 0.99% of slide height; #FFFFFF; center aligned]: NVLink
- photo at x 68.8, y 19.3, w 8.1, h 9.4 [#4A3F33]: Scale-Out tile photograph: dark metal Spectrum-X Ethernet switch chassis, three-quarter view with fan grille, warm light, cropped at edges (Boxed in a rounded-corner tile with thin grey stroke; image fills frame.)
- text at x 61.6, y 14.9, w 4.6, h 1.7 [NVIDIASans Medium, regular weight; font size 1.73% of slide height; #76B900; center aligned]: Scale-Out
- text at x 61.1, y 17, w 5.6, h 1 [NVIDIASans Regular, regular weight; font size 0.99% of slide height; #FFFFFF; center aligned]: Spectrum-X Ethernet
- illustration at x 78.3, y 20.1, w 7.4, h 8 [#000000]: Scale-Across tile: diagram of four grey switch boxes around a central box joined by thin white lines, tiny labels Scale-Up, Scale-Across, Scale-Out (Boxed in a rounded-corner tile with thin grey stroke; black background.)
- text at x 69.9, y 14.9, w 6, h 1.7 [NVIDIASans Medium, regular weight; font size 1.73% of slide height; #76B900; center aligned]: Scale-Across
- text at x 69.7, y 17, w 6.3, h 1 [NVIDIASans Regular, regular weight; font size 0.99% of slide height; #FFFFFF; center aligned]: Spectrum-XGS Ethernet
- photo at x 88, y 19.6, w 6.2, h 9.5 [#1C1C1C]: Scale-In tile photograph: dark circuit board with BlueField-4 DPU central chip, heatsinks and connectors, flat, on black (Boxed in a rounded-corner tile with thin grey stroke.)
- text at x 80.1, y 14.9, w 3.8, h 1.7 [NVIDIASans Medium, regular weight; font size 1.73% of slide height; #76B900; center aligned]: Scale-In
- text at x 79.8, y 17, w 4.2, h 1 [NVIDIASans Regular, regular weight; font size 0.99% of slide height; #FFFFFF; center aligned]: BlueField-4 DPU
- photo at x 53, y 19.3, w 3.5, h 9.4 [#2B2B2B]: Context Scale tile photograph: dark grey rack server chassis with open lid showing green-lit board and drive bays, front-top angle (Boxed in a rounded-corner tile with thin grey stroke.)
- text at x 87.7, y 14.9, w 6.4, h 1.7 [NVIDIASans Medium, regular weight; font size 1.73% of slide height; #76B900; center aligned]: Context Scale
- text at x 86.3, y 17, w 9.1, h 1 [NVIDIASans Regular, regular weight; font size 0.99% of slide height; #FFFFFF; center aligned]: Vera BlueField-4 Storage Processor
- shape at x 50.7, y 31, w 44.3, h 54.8 [#000000]: Chart panel: rounded rectangle outline enclosing the networking revenue line chart (No fill, thin grey stroke about #5A5A5A, roughly 1 px, corner radius about 1%.)
- text at x 61, y 34.3, w 23.7, h 2 [NVIDIASans Medium, regular weight; font size 1.98% of slide height; #FFFFFF; center aligned]: Networking Revenue (Indexed CY2020 = 100)
- chart at x 53, y 40.6, w 40, h 38 [#76B900]: Line chart: NVIDIA (green #76B900) and Total Market (grey #A1A1A1) indexed revenue CY2020 to CY2025, values labelled at every point, y axis 0 to 1,600 unitless, x axis years 2020 to 2025 (Legend centred above plot: green dash NVIDIA, grey dash Total Market, white text 1.48% of slide height. Y tick labels white at x 53.4% every 200 from 0 to 1,600, no gridlines. X baseline grey #8C8C8C at y about 76.5%, from x 56.9% to 91.9%. NVIDIA series: 100, 110, 169, 392, 594, 1435 (green labels). Total Market series: 100, 109, 139, 180, 218, 347 (grey labels). The 2025 NVIDIA point is highlighted with its label 1435 at top.)
- text at x 68.4, y 38.2, w 2.8, h 1.5 [NVIDIASans Regular, regular weight; font size 1.48% of slide height; #FFFFFF; left aligned]: NVIDIA
- text at x 72.9, y 38.2, w 5, h 1.5 [NVIDIASans Regular, regular weight; font size 1.48% of slide height; #FFFFFF; left aligned]: Total Market
- text at x 59.8, y 48, w 11.3, h 2 [NVIDIASans Medium, regular weight; font size 1.98% of slide height; #FFFFFF; left aligned]: CAGR (CY2020-2025)
- text at x 61.7, y 52.2, w 7.4, h 2 [NVIDIASans Medium, regular weight; font size 1.98% of slide height; #FFFFFF; left aligned]: NVIDIA – 70%
- text at x 61.7, y 56.5, w 7.4, h 2 [NVIDIASans Medium, regular weight; font size 1.98% of slide height; #FFFFFF; left aligned]: Market – 28%
- text at x 53.4, y 40.6, w 2.6, h 34.9 [NVIDIASans Regular, regular weight; font size 1.73% of slide height; #FFFFFF; right aligned]: 1,600 / 1,400 / 1,200 / 1,000 / 800 / 600 / 400 / 200 / 0 (y-axis tick labels)
- text at x 87.8, y 42.8, w 2.3, h 1.7 [NVIDIASans Medium, regular weight; font size 1.73% of slide height; #76B900; center aligned]: 1435
- text at x 82.2, y 58.9, w 1.7, h 1.7 [NVIDIASans Medium, regular weight; font size 1.73% of slide height; #76B900; center aligned]: 594
- text at x 76.4, y 65.3, w 1.7, h 1.7 [NVIDIASans Medium, regular weight; font size 1.73% of slide height; #76B900; center aligned]: 392
- text at x 88.1, y 66.7, w 1.7, h 1.7 [NVIDIASans Medium, regular weight; font size 1.73% of slide height; #A1A1A1; center aligned]: 347
- text at x 82.3, y 69.5, w 1.7, h 1.7 [NVIDIASans Medium, regular weight; font size 1.73% of slide height; #A1A1A1; center aligned]: 218
- text at x 70.6, y 68.4, w 1.7, h 1.7 [NVIDIASans Medium, regular weight; font size 1.73% of slide height; #76B900; center aligned]: 169
- text at x 76.4, y 70.4, w 1.7, h 1.7 [NVIDIASans Medium, regular weight; font size 1.73% of slide height; #A1A1A1; center aligned]: 180
- text at x 70.6, y 70.5, w 1.7, h 1.7 [NVIDIASans Medium, regular weight; font size 1.73% of slide height; #A1A1A1; center aligned]: 139
- text at x 64.7, y 69.7, w 1.7, h 1.7 [NVIDIASans Medium, regular weight; font size 1.73% of slide height; #76B900; center aligned]: 110
- text at x 64.8, y 72.1, w 1.7, h 1.7 [NVIDIASans Medium, regular weight; font size 1.73% of slide height; #A1A1A1; center aligned]: 109
- text at x 58.9, y 69.7, w 1.7, h 1.7 [NVIDIASans Medium, regular weight; font size 1.73% of slide height; #76B900; center aligned]: 100
- text at x 58.9, y 72.1, w 1.7, h 1.7 [NVIDIASans Medium, regular weight; font size 1.73% of slide height; #A1A1A1; center aligned]: 100
- text at x 58.6, y 78.1, w 31.5, h 1.7 [NVIDIASans Regular, regular weight; font size 1.73% of slide height; #FFFFFF; center aligned]: 2020 / 2021 / 2022 / 2023 / 2024 / 2025 (x-axis labels)
- logo at x 92.9, y 94.2, w 4.6, h 1.5 [#76B900]: NVIDIA logo: green eye mark with white NVIDIA wordmark (Bottom-right corner, no overlap.)

**Measured from the file** (exact, read from the PDF rather than estimated; x, y, w, h as % of the slide; font size as % of the slide height)

- fill at x 0.0, y 0.0, w 100.0, h 100.0: #000000 (covers the slide)
- image at x 59.7, y 18.9, w 8.2, h 10.6: 490 by 357 pixels
- image at x 68.8, y 19.3, w 8.1, h 9.4: 623 by 408 pixels
- image at x 88.0, y 19.6, w 6.2, h 9.5: 548 by 471 pixels
- image at x 78.3, y 20.1, w 7.4, h 8.0: 443 by 269 pixels
- image at x 53.0, y 19.3, w 3.5, h 9.4: 213 by 319 pixels
- text at x 5.2, y 22.2, w 34.5, h 4.9 [NVIDIASans-Medium; font size 4.94; #FFFFFF]: NVIDIA is the Largest and
- text at x 5.3, y 27.5, w 41.7, h 4.9 [NVIDIASans-Medium; font size 4.94; #FFFFFF]: Fastest Growing AI Networking
- text at x 5.3, y 32.8, w 29.1, h 4.9 [NVIDIASans-Medium; font size 4.94; #FFFFFF]: Company in the World
- text at x 52.8, y 14.9, w 4.2, h 1.7 [NVIDIASans-Medium; font size 1.73; #76B900]: Scale-Up
- text at x 61.6, y 14.9, w 4.6, h 1.7 [NVIDIASans-Medium; font size 1.73; #76B900]: Scale-Out
- text at x 69.9, y 14.9, w 6.0, h 1.7 [NVIDIASans-Medium; font size 1.73; #76B900]: Scale-Across
- text at x 80.1, y 14.9, w 3.8, h 1.7 [NVIDIASans-Medium; font size 1.73; #76B900]: Scale-In
- text at x 87.7, y 14.9, w 6.4, h 1.7 [NVIDIASans-Medium; font size 1.73; #76B900]: Context Scale
- text at x 54.0, y 17.0, w 1.8, h 1.0 [NVIDIASans-Regular; font size 0.99; #FFFFFF]: NVLink
- text at x 61.1, y 17.0, w 5.6, h 1.0 [NVIDIASans-Regular; font size 0.99; #FFFFFF]: Spectrum-X Ethernet
- text at x 69.7, y 17.0, w 6.3, h 1.0 [NVIDIASans-Regular; font size 0.99; #FFFFFF]: Spectrum-XGS Ethernet
- text at x 79.8, y 17.0, w 4.2, h 1.0 [NVIDIASans-Regular; font size 0.99; #FFFFFF]: BlueField-4 DPU
- text at x 86.3, y 17.0, w 9.1, h 1.0 [NVIDIASans-Regular; font size 0.99; #FFFFFF]: Vera BlueField-4 Storage Processor
- text at x 61.0, y 34.3, w 23.7, h 2.0 [NVIDIASans-Medium; font size 1.98; #FFFFFF]: Networking Revenue (Indexed CY2020 = 100)
- text at x 5.2, y 41.0, w 35.8, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: NVIDIA is the only company that offers five networking technologies:
- text at x 7.7, y 45.4, w 0.9, h 2.0 [NVIDIASans-Regular; font size 1.98; #76B900]: 1.
- text at x 9.1, y 45.4, w 9.3, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: Scale-Up (NVLink)
- text at x 53.4, y 40.6, w 2.6, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: 1,600
- text at x 59.8, y 48.0, w 11.3, h 2.0 [NVIDIASans-Medium; font size 1.98; #FFFFFF]: CAGR (CY2020-2025)​
- text at x 7.7, y 49.1, w 0.9, h 2.0 [NVIDIASans-Regular; font size 1.98; #76B900]: 2.
- text at x 9.1, y 49.1, w 23.1, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: Scale-Out (InfiniBand, Spectrum-X Ethernet)
- text at x 87.8, y 42.8, w 2.3, h 1.7 [NVIDIASans-Medium; font size 1.73; #76B900]: 1435
- text at x 53.4, y 45.0, w 2.6, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: 1,400
- text at x 68.4, y 38.2, w 2.8, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: NVIDIA
- text at x 72.9, y 38.2, w 5.0, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Total Market
- text at x 7.7, y 52.8, w 0.9, h 2.0 [NVIDIASans-Regular; font size 1.98; #76B900]: 3.
- text at x 9.1, y 52.8, w 20.3, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: Scale-Across (Spectrum-XGS Ethernet)
- text at x 61.7, y 52.2, w 7.4, h 2.0 [NVIDIASans-Medium; font size 1.98; #FFFFFF]: NVIDIA – 70% ​
- text at x 7.7, y 56.5, w 0.9, h 2.0 [NVIDIASans-Regular; font size 1.98; #76B900]: 4.
- text at x 9.1, y 56.5, w 33.1, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: Scale-In (BlueField-4 DPU and DOCA over Spectrum-X Ethernet)
- text at x 53.4, y 49.4, w 2.6, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: 1,200
- text at x 61.7, y 56.5, w 7.4, h 2.0 [NVIDIASans-Medium; font size 1.98; #FFFFFF]: Market – 28% ​
- text at x 7.7, y 60.2, w 0.9, h 2.0 [NVIDIASans-Regular; font size 1.98; #76B900]: 5.
- text at x 9.1, y 60.2, w 26.4, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: Context Scale (Vera BlueField-4 Storage Processor)
- text at x 53.4, y 53.8, w 2.6, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: 1,000
- text at x 5.3, y 64.3, w 37.2, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: Since the acquisition of Mellanox in 2020, NVIDIA’s Networking revenue
- text at x 5.3, y 66.7, w 36.6, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: has grown at a CAGR of 70%, materially outperforming industry peers.
- text at x 54.2, y 58.2, w 1.7, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: 800
- text at x 82.2, y 58.9, w 1.7, h 1.7 [NVIDIASans-Medium; font size 1.73; #76B900]: 594
- text at x 5.3, y 70.5, w 37.8, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: Today, NVIDIA is the largest and fastest growing AI Networking company
- text at x 54.2, y 62.6, w 1.7, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: 600
- text at x 5.3, y 72.9, w 6.7, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: in the world.
- text at x 76.4, y 65.3, w 1.7, h 1.7 [NVIDIASans-Medium; font size 1.73; #76B900]: 392
- text at x 88.1, y 66.7, w 1.7, h 1.7 [NVIDIASans-Medium; font size 1.73; #A1A1A1]: 347
- text at x 54.2, y 67.0, w 1.7, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: 400
- text at x 70.6, y 68.4, w 1.7, h 1.7 [NVIDIASans-Medium; font size 1.73; #76B900]: 169
- text at x 58.9, y 69.7, w 1.7, h 1.7 [NVIDIASans-Medium; font size 1.73; #76B900]: 100
- text at x 64.7, y 69.7, w 1.7, h 1.7 [NVIDIASans-Medium; font size 1.73; #76B900]: 110
- text at x 82.3, y 69.5, w 1.7, h 1.7 [NVIDIASans-Medium; font size 1.73; #A1A1A1]: 218
- text at x 70.6, y 70.5, w 1.7, h 1.7 [NVIDIASans-Medium; font size 1.73; #A1A1A1]: 139
- text at x 76.4, y 70.4, w 1.7, h 1.7 [NVIDIASans-Medium; font size 1.73; #A1A1A1]: 180
- text at x 54.2, y 71.4, w 1.7, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: 200
- text at x 58.9, y 72.1, w 1.7, h 1.7 [NVIDIASans-Medium; font size 1.73; #A1A1A1]: 100
- text at x 64.8, y 72.1, w 1.7, h 1.7 [NVIDIASans-Medium; font size 1.73; #A1A1A1]: 109
- text at x 55.3, y 75.7, w 0.6, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: 0
- text at x 58.6, y 78.1, w 2.3, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: 2020
- text at x 64.4, y 78.1, w 2.3, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: 2021
- text at x 70.3, y 78.1, w 2.3, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: 2022
- text at x 76.1, y 78.1, w 2.3, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: 2023
- text at x 82.0, y 78.1, w 2.3, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: 2024
- text at x 87.8, y 78.1, w 2.3, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: 2025

### Slide 18

**Review notes**

- Review: Weaknesses are consequential: many slides depend on small gray body text and tiny labels (slides 4, 8, 13, 18, 21, 22), the financial tables on slide 9 are unreadable at presentation size, and slide 2 is a dense disclaimer wall.
- Suggested repair: Enlarge or simplify the small body text and diagram labels on slides 4, 8, 13, 18 and 21, and replace the screenshot-style tables on slide 9 with a few legible headline figures.

**Text**

NVLink Fusion Expands NVIDIA Opportunities
NVIDIA NVLink Fusion offers customers the option to build semi-custom AI factories by integrating their custom XPUs or CPUs with NVIDIA’s proven rack architecture.
And to accelerate adoption, NVIDIA is partnering with various CPU, custom silicon, EDA and optical interconnect suppliers.
Customer reception has been strong with AWS, Intel, RIKEN, and others already committed to adopting NVLink scale-up technology.
NVLink Fusion opportunity to approach $250 billion by the end of the decade.
NVLink Connectors
XPU
XPU Package
NVIDIA NVLink 6
Switch
NVIDIA Vera or
Custom CPU
NVIDIA NVLink
Cable Cartridge
NVIDIA Grace
BlueField
NVIDIA ConnectX-9
Compute Tray x18
NVIDIA NVLink Switch Tray x9
NVIDIA NVLink Spine x1
NVIDIA MGX Rack
Cableless, Hoseless, Modular Design
3.6 TB/s per XPU All-to-All Bandwidth
Connects 72 XPUs in a Single Domain
Third-Generation Design
45C Inlet Temp
3X Lower Latency
100% Liquid Cooled
Built-In Factory Resiliency

**Layout**

Centred single-column header with the title at the top and four centred body lines beneath it, followed by a horizontal row of one XPU chip and four product columns (Compute Tray, NVLink Switch Tray, NVLink Spine, MGX Rack), each with a centred bold caption and small centred sub-lines, joined left to right by two white arrows. Left margin of the diagram row is about 6.7% and right margin about 17% for the rack; the NVIDIA logo sits bottom right. No footer text and no slide number.

**Visuals**

Five visuals on a black field. (1) XPU: a square dark-grey (#262626) chip package about 8.5% by 15.5% of the slide with tan (#C8B98A) contact pads along its top, bottom, left and right edges, and a blue (#0074DF, 90% opacity) die with a grid of lighter blue cells labelled XPU in white at its centre. (2) White arrow pointing right from the XPU to the Compute Tray. (3) Photograph of a Compute Tray x18: a black-background, backlit, line-drawn style top view of a server tray with four glowing XPU packages in a row at the top, two smaller glowing Vera or custom CPU modules below them, a single glowing Grace BlueField module lower centre and two small ConnectX-9 modules at the bottom; lit with amber and white highlights, full frame, boxed to about 10.9% by 34.1%, dominant colours black, gold and amber. Leader lines and labels: NVLink Connectors (white, from the top edge of the tray diagonally up-right to the label), XPU Package (horizontal line to the upper-row packages), NVIDIA Vera or Custom CPU (horizontal line to a CPU module), NVIDIA Grace BlueField (horizontal line to the Grace module), NVIDIA ConnectX-9 (horizontal line to the bottom modules). (4) Photograph of an NVIDIA NVLink Switch Tray x9: a black-background, backlit line-drawn view of a server tray with four glowing rectangular NVLink 6 switch chips arranged in a V shape at the upper centre, with cabling and board detail below, boxed to about 10.9% by 34.0%, dominant colours black, gold and amber. Leader lines: NVIDIA NVLink 6 Switch (horizontal white line to the right-hand switch chips) and NVLink Connectors (diagonal white line from the top connector area of this tray to the label above). (5) Photograph of an NVIDIA NVLink Spine x1: four tall black vertical cable-cartridge modules standing side by side, each with a column of small glowing gold/amber contact points down its inner edge, on a black background, boxed to about 11.3% by 28.7%; a white leader line from the NVIDIA NVLink Cable Cartridge label points to the left-hand module. (6) White arrow pointing right from the Spine to the MGX Rack. (7) Photograph of an NVIDIA MGX Rack: a tall black enclosure with a thin grey frame, a brushed-gold lit band of horizontal rails across its middle, dark top and base, shown front-on and centred, boxed to about 6.6% by 44.9%, dominant colours black and gold. (8) NVIDIA logo bottom right in green and white. Nothing is cropped, full-bleed or tinted except the photographs which sit on the black background.

**Design**

Background: Solid black #000000 covering the whole slide. Black (#000000) background throughout with no gradient or texture. Title in white (#FFFFFF) NVIDIA Sans Medium (a geometric grotesque sans-serif), about 4.9% of slide height, centred. Body paragraphs in mid-grey (#A1A1A1) NVIDIA Sans Regular, about 2.0% of slide height, centred, with generous line spacing. Sub-captions and labels in white (#FFFFFF) NVIDIA Sans Regular about 1.5% (labels) and 1.2% (sub-lines); column headings in white NVIDIA Sans Medium about 1.7%. Accent colours: XPU die blue #0074DF at 90% opacity, chip frame #262626, tan pads #C8B98A, NVIDIA logo green #76B900. Leader lines and arrows are thin white strokes (about 1 px) with no shadows, no rounded corners and no gradients.

**Elements and positions** (back to front; x, y, width, height as % of the slide; type, colour and alignment in brackets)

- shape at x 0, y 0, w 100, h 100 [#000000]: Full-slide background rectangle (Solid fill, no border, covers the slide)
- illustration at x 6.7, y 38.7, w 8.5, h 15.5 [#262626]: XPU chip package with tan contact pads and blue die (Dark grey square body, tan #C8B98A pads on all four edges, blue #0074DF die (see separate shape) at 90% opacity overlaid on its centre)
- shape at x 8.4, y 41.9, w 5.1, h 9.1 [#0074DF]: XPU die, blue with label XPU (Filled rectangle, 90% opacity, grid of lighter blue cells, no border)
- text at x 10, y 45.7, w 2, h 1.7 [NVIDIA Sans Medium; font size 1.73% of slide height; #FFFFFF; center aligned]: XPU
- photo at x 20.8, y 40, w 10.9, h 34.1: Compute Tray x18 photograph (Backlit line-drawn server tray on black, full frame, boxed, no overlay)
- photo at x 43.8, y 39.9, w 10.9, h 34: NVLink Switch Tray x9 photograph (Backlit line-drawn switch tray on black, full frame, boxed, no overlay)
- photo at x 64.8, y 42.6, w 11.3, h 28.7: NVLink Spine x1 photograph (Four black cable-cartridge modules with glowing gold contacts on black, no overlay)
- photo at x 83.4, y 34.5, w 6.6, h 44.9: NVIDIA MGX Rack photograph (Front view of a black rack with a lit gold rail band, on black, no overlay)
- line at x 16.2, y 46.2, w 4.6, h 0.3 [#FFFFFF]: Arrow from XPU to Compute Tray (White 1 px line with filled triangular arrowhead at its right end)
- line at x 77.8, y 56.6, w 4.6, h 0.3 [#FFFFFF]: Arrow from NVLink Spine to MGX Rack (White 1 px line with filled triangular arrowhead at its right end)
- line at x 29.9, y 37.5, w 4.4, h 5.3 [#FFFFFF]: Leader line to NVLink Connectors on Compute Tray (Thin white diagonal from tray top-left to label)
- line at x 42.5, y 37.3, w 5, h 6.3 [#FFFFFF]: Leader line to NVLink Connectors on Switch Tray (Thin white diagonal from tray top to label)
- line at x 30, y 46.4, w 4.2, h 0.1 [#FFFFFF]: Leader line to XPU Package (Thin white horizontal line)
- line at x 30, y 51.2, w 4.2, h 0.1 [#FFFFFF]: Leader line to NVIDIA Vera or Custom CPU (Thin white horizontal line)
- line at x 51.8, y 47.6, w 3.8, h 0.1 [#FFFFFF]: Leader line to NVIDIA NVLink 6 Switch (Thin white horizontal line)
- line at x 63.1, y 56.9, w 2.2, h 0.1 [#FFFFFF]: Leader line to NVIDIA NVLink Cable Cartridge (Thin white horizontal line)
- line at x 26, y 62.2, w 8.2, h 0.1 [#FFFFFF]: Leader line to NVIDIA Grace BlueField (Thin white horizontal line)
- line at x 30, y 66.7, w 4.2, h 0.1 [#FFFFFF]: Leader line to NVIDIA ConnectX-9 (Thin white horizontal line)
- text at x 34.8, y 36.5, w 7.5, h 1.5 [NVIDIA Sans Regular; font size 1.48% of slide height; #FFFFFF; left aligned]: NVLink Connectors
- text at x 34.9, y 45.8, w 5.2, h 1.5 [NVIDIA Sans Regular; font size 1.48% of slide height; #FFFFFF; left aligned]: XPU Package
- text at x 34.8, y 50.5, w 5.9, h 3.3 [NVIDIA Sans Regular; font size 1.48% of slide height; #FFFFFF; left aligned]: NVIDIA Vera or / Custom CPU (Two lines)
- text at x 56.1, y 47.1, w 6.7, h 3.2 [NVIDIA Sans Regular; font size 1.48% of slide height; #FFFFFF; left aligned]: NVIDIA NVLink 6 / Switch (Two lines)
- text at x 56.4, y 56.2, w 6.1, h 3.3 [NVIDIA Sans Regular; font size 1.48% of slide height; #FFFFFF; right aligned]: NVIDIA NVLink / Cable Cartridge (Two lines, right-aligned)
- text at x 34.8, y 61.7, w 5.4, h 3.3 [NVIDIA Sans Regular; font size 1.48% of slide height; #FFFFFF; left aligned]: NVIDIA Grace / BlueField (Two lines)
- text at x 34.8, y 66.1, w 7.6, h 1.5 [NVIDIA Sans Regular; font size 1.48% of slide height; #FFFFFF; left aligned]: NVIDIA ConnectX-9
- text at x 42.4, y 80.6, w 13.8, h 1.7 [NVIDIA Sans Medium; font size 1.73% of slide height; #FFFFFF; center aligned]: NVIDIA NVLink Switch Tray x9 heading
- text at x 22.1, y 80.6, w 8.3, h 1.7 [NVIDIA Sans Medium; font size 1.73% of slide height; #FFFFFF; center aligned]: Compute Tray x18
- text at x 65.5, y 80.6, w 11, h 1.7 [NVIDIA Sans Medium; font size 1.73% of slide height; #FFFFFF; center aligned]: NVIDIA NVLink Spine x1
- text at x 82.5, y 80.6, w 8.3, h 1.7 [NVIDIA Sans Medium; font size 1.73% of slide height; #FFFFFF; center aligned]: NVIDIA MGX Rack
- text at x 20.4, y 83.2, w 11.7, h 3 [NVIDIA Sans Regular; font size 1.24% of slide height; #FFFFFF; center aligned]: Cableless, Hoseless, Modular Design / 45C Inlet Temp (Two lines)
- text at x 43.2, y 83.2, w 12.1, h 4.6 [NVIDIA Sans Regular; font size 1.24% of slide height; #FFFFFF; center aligned]: 3.6 TB/s per XPU All-to-All Bandwidth / 3X Lower Latency / Built-In Factory Resiliency (Three lines)
- text at x 64.9, y 83.2, w 12.2, h 1.2 [NVIDIA Sans Regular; font size 1.24% of slide height; #FFFFFF; center aligned]: Connects 72 XPUs in a Single Domain
- text at x 82.7, y 83.2, w 7.9, h 3 [NVIDIA Sans Regular; font size 1.24% of slide height; #FFFFFF; center aligned]: Third-Generation Design / 100% Liquid Cooled (Two lines)
- text at x 20.1, y 10.5, w 59.7, h 4.9 [NVIDIA Sans Medium; font size 4.94% of slide height; #FFFFFF; center aligned]: NVLink Fusion Expands NVIDIA Opportunities
- text at x 6.9, y 18.1, w 86.6, h 4 [NVIDIA Sans Regular; font size 1.98% of slide height; #A1A1A1; center aligned]: NVIDIA NVLink Fusion offers customers the option to build semi-custom AI factories by integrating their custom XPUs or CPUs with NVIDIA’s proven rack architecture. / And to accelerate adoption, NVIDIA is partnering with various CPU, custom silicon, EDA and optical interconnect suppliers. (Two lines)
- text at x 15.9, y 24.8, w 68.6, h 2 [NVIDIA Sans Regular; font size 1.98% of slide height; #A1A1A1; center aligned]: Customer reception has been strong with AWS, Intel, RIKEN, and others already committed to adopting NVLink scale-up technology.
- text at x 30, y 28.9, w 40.3, h 2 [NVIDIA Sans Regular; font size 1.98% of slide height; #A1A1A1; center aligned]: NVLink Fusion opportunity to approach $250 billion by the end of the decade.
- logo at x 92.9, y 94.2, w 4.6, h 1.6 [#76B900]: NVIDIA logo, green mark with white wordmark (Bottom-right corner logo)

**Measured from the file** (exact, read from the PDF rather than estimated; x, y, w, h as % of the slide; font size as % of the slide height)

- fill at x 0.0, y 0.0, w 100.0, h 100.0: #000000 (covers the slide)
- fill at x 6.7, y 38.7, w 8.5, h 15.5: #262626
- fill at x 8.4, y 41.9, w 5.1, h 9.1: #0074DF, opacity 0.90
- image at x 43.8, y 39.9, w 10.9, h 34.0: 404 by 710 pixels
- image at x 20.8, y 40.0, w 10.9, h 34.1: 403 by 711 pixels
- image at x 64.8, y 42.6, w 11.3, h 28.7: 646 by 923 pixels
- image at x 83.4, y 34.5, w 6.6, h 44.9: 470 by 1797 pixels
- text at x 20.1, y 10.5, w 59.7, h 4.9 [NVIDIASans-Medium; font size 4.94; #FFFFFF]: NVLink Fusion Expands NVIDIA Opportunities
- text at x 6.9, y 18.1, w 86.6, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: NVIDIA NVLink Fusion offers customers the option to build semi-custom AI factories by integrating their custom XPUs or CPUs with NVIDIA’s proven rack architecture.
- text at x 18.5, y 20.7, w 63.3, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: And to accelerate adoption, NVIDIA is partnering with various CPU, custom silicon, EDA and optical interconnect suppliers.
- text at x 15.9, y 24.8, w 68.6, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: Customer reception has been strong with AWS, Intel, RIKEN, and others already committed to adopting NVLink scale-up technology.
- text at x 30.0, y 28.9, w 40.3, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: NVLink Fusion opportunity to approach $250 billion by the end of the decade.
- text at x 34.8, y 36.5, w 7.5, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: NVLink Connectors
- text at x 10.0, y 45.7, w 2.0, h 1.7 [NVIDIASans-Medium; font size 1.73; #FFFFFF]: XPU
- text at x 34.9, y 45.8, w 5.2, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: XPU Package
- text at x 56.1, y 47.1, w 6.7, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: NVIDIA NVLink 6
- text at x 56.1, y 48.8, w 2.7, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Switch
- text at x 34.8, y 50.5, w 5.9, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: NVIDIA Vera or
- text at x 34.8, y 52.3, w 5.0, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Custom CPU
- text at x 56.7, y 56.2, w 5.7, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: NVIDIA NVLink
- text at x 56.4, y 58.0, w 6.1, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Cable Cartridge
- text at x 34.8, y 61.7, w 5.4, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: NVIDIA Grace
- text at x 34.8, y 63.5, w 3.5, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: BlueField
- text at x 34.8, y 66.1, w 7.6, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: NVIDIA ConnectX-9
- text at x 22.1, y 80.6, w 8.3, h 1.7 [NVIDIASans-Medium; font size 1.73; #FFFFFF]: Compute Tray x18
- text at x 42.4, y 80.6, w 13.8, h 1.7 [NVIDIASans-Medium; font size 1.73; #FFFFFF]: NVIDIA NVLink Switch Tray x9
- text at x 65.5, y 80.6, w 11.0, h 1.7 [NVIDIASans-Medium; font size 1.73; #FFFFFF]: NVIDIA NVLink Spine x1
- text at x 82.5, y 80.6, w 8.3, h 1.7 [NVIDIASans-Medium; font size 1.73; #FFFFFF]: NVIDIA MGX Rack
- text at x 20.4, y 83.2, w 11.7, h 1.2 [NVIDIASans-Regular; font size 1.24; #FFFFFF]: Cableless, Hoseless, Modular Design
- text at x 43.2, y 83.2, w 12.1, h 1.2 [NVIDIASans-Regular; font size 1.24; #FFFFFF]: 3.6 TB/s per XPU All-to-All Bandwidth
- text at x 64.9, y 83.2, w 12.2, h 1.2 [NVIDIASans-Regular; font size 1.24; #FFFFFF]: Connects 72 XPUs in a Single Domain
- text at x 82.7, y 83.2, w 7.9, h 1.2 [NVIDIASans-Regular; font size 1.24; #FFFFFF]: Third-Generation Design
- text at x 23.8, y 84.9, w 4.9, h 1.2 [NVIDIASans-Regular; font size 1.24; #FFFFFF]: 45C Inlet Temp
- text at x 46.4, y 84.9, w 5.8, h 1.2 [NVIDIASans-Regular; font size 1.24; #FFFFFF]: 3X Lower Latency
- text at x 83.5, y 84.9, w 6.4, h 1.2 [NVIDIASans-Regular; font size 1.24; #FFFFFF]: 100% Liquid Cooled
- text at x 45.1, y 86.6, w 8.4, h 1.2 [NVIDIASans-Regular; font size 1.24; #FFFFFF]: Built-In Factory Resiliency

### Slide 19

**Text**

NVIDIA CUDA Turns Programmability into Ecosystem Scale
NVIDIA CUDA defines NVIDIA’s strategy and powers the flywheel. Developers drive new applications. New applications create new ecosystems.
New ecosystems grow the installed base. And a larger installed base invites more developers.
NVIDIA compute accelerates every workload. AI and non-AI. All fields of science.
NVIDIA continuously updates its software, driving down the cost of computing, and improving the installed base’s productivity.
Built on CUDA, NVIDIA compute is versatile, fungible, durable, and therefore investable.
NEW
APPLICATIONS
ECOSYSTEMS
DEVELOPERS
INSTALLED
BASE
20 YEARS
CUDA
FLYWHEEL
Developers
3M
11M
2021
2026 Q2
AI
Startups
9K
50K
2021
2026 Q2
CUDA Downloads
32M
1.2B
2021
2026 Q2
GPU -Accelerated
Applications
2K
9K
2021
2026 Q2
NVIDIA

**Layout**

Single black slide with a centred title at the top and a centred four-paragraph subtitle block beneath it. The lower half is split into two zones: a circular flywheel diagram on the left and a 2×2 grid of bubble charts on the right, each chart with its label at left, a thin baseline, and two year columns (2021 and 2026 Q2). The NVIDIA logo sits in the bottom-right corner. No footer or slide number is visible.

**Visuals**

Left: a circular flywheel diagram with four white hand-drawn curved arrows arranged clockwise in a ring around a green handwritten label. The ring's four nodes are NEW APPLICATIONS (top), DEVELOPERS (right), INSTALLED BASE (bottom) and ECOSYSTEMS (left); the arrows run ECOSYSTEMS to NEW APPLICATIONS (arrowhead at the top), NEW APPLICATIONS to DEVELOPERS (arrowhead at the upper right), DEVELOPERS to INSTALLED BASE (arrowhead at the lower right), and INSTALLED BASE to ECOSYSTEMS (arrowhead at the lower left). The centre reads 20 YEARS / CUDA / FLYWHEEL in #76B900 handwritten capitals. Right: four bubble charts, each with two solid #76B900 circles scaled by value. Developers: 3M (2021, small circle) and 11M (2026 Q2, large circle), circle sizes about proportional to value. AI Startups: 9K (2021) and 50K (2026 Q2). CUDA Downloads: 32M (2021, smallest circle) and 1.2B (2026 Q2, largest circle). GPU-Accelerated Applications: 2K (2021) and 9K (2026 Q2). Each chart has a thin grey baseline with the category label to its left. Values are printed above the circles in bold white; no axes or legend. Only the 2026 Q2 values are printed larger than the 2021 values in the same chart; the circle areas are not labelled as estimates.

**Design**

Background: Solid black #000000 covering the whole slide. Black background (#000000) with white titles (#FFFFFF), grey body text (#A1A1A1) and bright green accents (#76B900). Typeface is NVIDIA Sans (a geometric grotesque sans-serif) in Medium for the title, Regular for body and labels, and Bold for the bubble values. Title is about 4.9% of slide height; subtitle and chart labels about 2.0%; bubble values and year labels about 1.7%. Lines are thin grey hairlines about 0.1% high. Bubbles are flat solid green circles with no outline, shadow or gradient. Arrows are white with rough hand-drawn edges.

**Elements and positions** (back to front; x, y, width, height as % of the slide; type, colour and alignment in brackets)

- shape at x 0, y 0, w 100, h 100 [#000000]: full-slide black background fill
- illustration at x 8.2, y 43.2, w 8.2, h 17.8 [#FFFFFF]: flywheel arrow, ECOSYSTEMS to NEW APPLICATIONS, arrowhead at top (hand-drawn white arc with arrowhead at lower-left end, transparent background)
- illustration at x 27.1, y 43.2, w 8.4, h 17.8 [#FFFFFF]: flywheel arrow, NEW APPLICATIONS to DEVELOPERS, arrowhead at top right (hand-drawn white arc with arrowhead at upper-right end, transparent background)
- illustration at x 8.2, y 67.2, w 9.5, h 19.5 [#FFFFFF]: flywheel arrow, ECOSYSTEMS to INSTALLED BASE, arrowhead at bottom (hand-drawn white arc with arrowhead at lower-right end, transparent background)
- illustration at x 25.1, y 66.1, w 10.4, h 19.6 [#FFFFFF]: flywheel arrow, DEVELOPERS to INSTALLED BASE, arrowhead at bottom left (hand-drawn white arc with arrowhead at lower-left end, transparent background)
- illustration at x 16.6, y 57.1, w 10.4, h 12.8 [#76B900]: centre label 20 YEARS CUDA FLYWHEEL in green handwritten capitals (three-line green handwritten lettering, transparent background)
- illustration at x 17.3, y 40.4, w 8.9, h 4.8 [#FFFFFF]: flywheel node label NEW APPLICATIONS, white thin sans capitals on two lines (rendered as image, centre aligned)
- illustration at x 5.2, y 62.5, w 8.2, h 1.9 [#FFFFFF]: flywheel node label ECOSYSTEMS, white thin sans capitals (rendered as image, right of arrow)
- illustration at x 30.8, y 62.5, w 8, h 1.9 [#FFFFFF]: flywheel node label DEVELOPERS, white thin sans capitals (rendered as image)
- illustration at x 18.5, y 82.9, w 6.6, h 4.8 [#FFFFFF]: flywheel node label INSTALLED BASE, white thin sans capitals on two lines (rendered as image, centre aligned)
- title at x 11.5, y 10.5, w 76.9, h 4.9 [NVIDIA Sans Medium, sans-serif, regular case; font size 4.94% of slide height; #FFFFFF; center aligned]: NVIDIA CUDA Turns Programmability into Ecosystem Scale
- body at x 13.2, y 18.1, w 73.9, h 2 [NVIDIA Sans Regular, sans-serif; font size 1.98% of slide height; #A1A1A1; center aligned]: NVIDIA CUDA defines NVIDIA’s strategy and powers the flywheel. Developers drive new applications. New applications create new ecosystems.
- body at x 26, y 20.7, w 48.3, h 2 [NVIDIA Sans Regular, sans-serif; font size 1.98% of slide height; #A1A1A1; center aligned]: New ecosystems grow the installed base. And a larger installed base invites more developers.
- body at x 29.4, y 24.8, w 41.5, h 2 [NVIDIA Sans Regular, sans-serif; font size 1.98% of slide height; #A1A1A1; center aligned]: NVIDIA compute accelerates every workload. AI and non-AI. All fields of science.
- body at x 17.3, y 28.9, w 65.7, h 2 [NVIDIA Sans Regular, sans-serif; font size 1.98% of slide height; #A1A1A1; center aligned]: NVIDIA continuously updates its software, driving down the cost of computing, and improving the installed base’s productivity.
- body at x 27.6, y 33, w 44.8, h 2 [NVIDIA Sans Regular, sans-serif; font size 1.98% of slide height; #A1A1A1; center aligned]: Built on CUDA, NVIDIA compute is versatile, fungible, durable, and therefore investable.
- line at x 50.9, y 65, w 12.9, h 0.1 [#8C8C8C]: baseline under Developers chart (hairline, stroke about 0.1% of slide height)
- shape at x 51.05, y 47.6, w 3.25, h 5.8 [#76B900]: bubble 3M, 2021 Developers (solid circle, no outline)
- shape at x 56.8, y 44.9, w 6.3, h 11.2 [#76B900]: bubble 11M, 2026 Q2 Developers (solid circle, no outline)
- text at x 44.1, y 57.1, w 5.8, h 2 [NVIDIA Sans Regular, sans-serif; font size 1.98% of slide height; #FFFFFF; right aligned]: Developers
- text at x 51.9, y 45.3, w 1.5, h 1.7 [NVIDIA Sans Bold, sans-serif; font size 1.73% of slide height; #FFFFFF; center aligned]: 3M
- text at x 58.9, y 42.6, w 2.1, h 1.7 [NVIDIA Sans Bold, sans-serif; font size 1.73% of slide height; #FFFFFF; center aligned]: 11M
- text at x 52.1, y 59.2, w 10.2, h 1.7 [NVIDIA Sans Regular, sans-serif; font size 1.73% of slide height; #FFFFFF; left aligned]: 2021 2026 Q2 (year labels 2021 under small bubble and 2026 Q2 under large bubble)
- line at x 77.4, y 65, w 13, h 0.1 [#8C8C8C]: baseline under AI Startups chart (hairline)
- shape at x 77.1, y 48.1, w 2.65, h 4.6 [#76B900]: bubble 9K, 2021 AI Startups (solid circle, no outline)
- shape at x 82.5, y 44.8, w 6.3, h 11.2 [#76B900]: bubble 50K, 2026 Q2 AI Startups (solid circle, no outline)
- text at x 70.2, y 57, w 1.5, h 2.2 [NVIDIA Sans Regular, sans-serif; font size 2.22% of slide height; #FFFFFF; right aligned]: AI
- text at x 71.7, y 57.2, w 4.6, h 2 [NVIDIA Sans Regular, sans-serif; font size 1.98% of slide height; #FFFFFF; right aligned]: Startups
- text at x 77.8, y 45.8, w 1.2, h 1.7 [NVIDIA Sans Bold, sans-serif; font size 1.73% of slide height; #FFFFFF; center aligned]: 9K
- text at x 84.8, y 42.6, w 1.8, h 1.7 [NVIDIA Sans Bold, sans-serif; font size 1.73% of slide height; #FFFFFF; center aligned]: 50K
- text at x 78, y 59.2, w 10.2, h 1.7 [NVIDIA Sans Regular, sans-serif; font size 1.73% of slide height; #FFFFFF; left aligned]: 2021 2026 Q2 (year labels 2021 and 2026 Q2 under the AI Startups bubbles)
- line at x 51.3, y 83.6, w 13.1, h 0.1 [#8C8C8C]: baseline under CUDA Downloads chart (hairline)
- shape at x 52.7, y 74.9, w 1, h 1.8 [#76B900]: bubble 32M, 2021 CUDA Downloads (solid circle, no outline)
- shape at x 57.3, y 70.2, w 6.3, h 11.2 [#76B900]: bubble 1.2B, 2026 Q2 CUDA Downloads (solid circle, no outline)
- text at x 40.8, y 82.8, w 9, h 2 [NVIDIA Sans Regular, sans-serif; font size 1.98% of slide height; #FFFFFF; right aligned]: CUDA Downloads
- text at x 52.2, y 72.6, w 2.1, h 1.7 [NVIDIA Sans Bold, sans-serif; font size 1.73% of slide height; #FFFFFF; center aligned]: 32M
- text at x 59.4, y 67.9, w 2.1, h 1.7 [NVIDIA Sans Bold, sans-serif; font size 1.73% of slide height; #FFFFFF; center aligned]: 1.2B
- text at x 52.1, y 84.9, w 10.2, h 1.7 [NVIDIA Sans Regular, sans-serif; font size 1.73% of slide height; #FFFFFF; left aligned]: 2021 2026 Q2 (year labels 2021 and 2026 Q2 under the CUDA Downloads bubbles)
- line at x 77.7, y 83.6, w 13, h 0.1 [#8C8C8C]: baseline under GPU-Accelerated Applications chart (hairline)
- shape at x 76.95, y 73.2, w 2.95, h 5.2 [#76B900]: bubble 2K, 2021 GPU-Accelerated Applications (solid circle, no outline)
- shape at x 82.5, y 70.2, w 6.3, h 11.2 [#76B900]: bubble 9K, 2026 Q2 GPU-Accelerated Applications (solid circle, no outline)
- text at x 67.2, y 81.2, w 9.4, h 2 [NVIDIA Sans Regular, sans-serif; font size 1.98% of slide height; #FFFFFF; right aligned]: GPU -Accelerated Applications
- text at x 77.8, y 70.8, w 1.2, h 1.7 [NVIDIA Sans Bold, sans-serif; font size 1.73% of slide height; #FFFFFF; center aligned]: 2K
- text at x 85.1, y 67.9, w 1.2, h 1.7 [NVIDIA Sans Bold, sans-serif; font size 1.73% of slide height; #FFFFFF; center aligned]: 9K
- text at x 78, y 84.9, w 10.2, h 1.7 [NVIDIA Sans Regular, sans-serif; font size 1.73% of slide height; #FFFFFF; left aligned]: 2021 2026 Q2 (year labels 2021 and 2026 Q2 under the GPU-Accelerated Applications bubbles)
- logo at x 92.9, y 94.2, w 4.6, h 1.6 [#76B900]: NVIDIA logo: green eye mark with white NVIDIA wordmark (bottom-right corner)

**Measured from the file** (exact, read from the PDF rather than estimated; x, y, w, h as % of the slide; font size as % of the slide height)

- fill at x 0.0, y 0.0, w 100.0, h 100.0: #000000 (covers the slide)
- fill at x 57.3, y 70.2, w 6.3, h 11.2: #76B900
- fill at x 56.8, y 44.9, w 6.3, h 11.2: #76B900
- fill at x 82.5, y 44.8, w 6.3, h 11.2: #76B900
- fill at x 82.5, y 70.2, w 6.3, h 11.2: #76B900
- image at x 25.1, y 66.1, w 10.4, h 19.6: 584 by 616 pixels
- image at x 8.2, y 67.2, w 9.5, h 19.5: 532 by 614 pixels
- image at x 27.1, y 43.2, w 8.4, h 17.8: 472 by 561 pixels
- image at x 8.2, y 43.2, w 8.2, h 17.8: 460 by 561 pixels
- image at x 16.6, y 57.1, w 10.4, h 12.8: 466 by 321 pixels
- image at x 17.3, y 40.4, w 8.9, h 4.8: 652 by 197 pixels
- image at x 18.5, y 82.9, w 6.6, h 4.8: 479 by 196 pixels
- image at x 5.2, y 62.5, w 8.2, h 1.9: 598 by 77 pixels
- image at x 30.8, y 62.5, w 8.0, h 1.9: 583 by 77 pixels
- text at x 11.5, y 10.5, w 76.9, h 4.9 [NVIDIASans-Medium; font size 4.94; #FFFFFF]: NVIDIA CUDA Turns Programmability into Ecosystem Scale
- text at x 13.2, y 18.1, w 73.9, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: NVIDIA CUDA defines NVIDIA’s strategy and powers the flywheel. Developers drive new applications. New applications create new ecosystems.
- text at x 26.0, y 20.7, w 48.3, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: New ecosystems grow the installed base. And a larger installed base invites more developers.
- text at x 29.4, y 24.8, w 41.5, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: NVIDIA compute accelerates every workload. AI and non-AI. All fields of science.
- text at x 17.3, y 28.9, w 65.7, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: NVIDIA continuously updates its software, driving down the cost of computing, and improving the installed base’s productivity.
- text at x 27.6, y 33.0, w 44.8, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: Built on CUDA, NVIDIA compute is versatile, fungible, durable, and therefore investable.
- text at x 58.9, y 42.6, w 2.1, h 1.7 [NVIDIASans-Bold; font size 1.73; #FFFFFF]: 11M
- text at x 84.8, y 42.6, w 1.8, h 1.7 [NVIDIASans-Bold; font size 1.73; #FFFFFF]: 50K
- text at x 70.2, y 57.0, w 1.5, h 2.2 [NVIDIASans-Regular; font size 2.22; #FFFFFF]: AI
- text at x 51.9, y 45.3, w 1.5, h 1.7 [NVIDIASans-Bold; font size 1.73; #FFFFFF]: 3M
- text at x 77.8, y 45.8, w 1.2, h 1.7 [NVIDIASans-Bold; font size 1.73; #FFFFFF]: 9K
- text at x 44.1, y 57.1, w 5.8, h 2.0 [NVIDIASans-Regular; font size 1.98; #FFFFFF]: Developers
- text at x 71.7, y 57.2, w 4.6, h 2.0 [NVIDIASans-Regular; font size 1.98; #FFFFFF]: Startups
- text at x 52.1, y 59.2, w 10.2, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: 2021 2026 Q2
- text at x 78.0, y 59.2, w 10.2, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: 2021 2026 Q2
- text at x 59.4, y 67.9, w 2.1, h 1.7 [NVIDIASans-Bold; font size 1.73; #FFFFFF]: 1.2B
- text at x 85.1, y 67.9, w 1.2, h 1.7 [NVIDIASans-Bold; font size 1.73; #FFFFFF]: 9K
- text at x 67.2, y 81.2, w 9.4, h 2.0 [NVIDIASans-Regular; font size 1.98; #FFFFFF]: GPU -Accelerated
- text at x 77.8, y 70.8, w 1.2, h 1.7 [NVIDIASans-Bold; font size 1.73; #FFFFFF]: 2K
- text at x 40.8, y 82.8, w 9.0, h 2.0 [NVIDIASans-Regular; font size 1.98; #FFFFFF]: CUDA Downloads
- text at x 52.2, y 72.6, w 2.1, h 1.7 [NVIDIASans-Bold; font size 1.73; #FFFFFF]: 32M
- text at x 69.9, y 83.5, w 6.4, h 2.0 [NVIDIASans-Regular; font size 1.98; #FFFFFF]: Applications
- text at x 52.1, y 84.9, w 10.2, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: 2021 2026 Q2
- text at x 78.0, y 84.9, w 10.2, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: 2021 2026 Q2

### Slide 20

**Text**

NVIDIA CUDA-X Accelerates
Apps and Reduces the Cost of Computing
CUDA-X libraries accelerate apps, enabling them to run orders of magnitude faster, more efficiently, and at lower cost than on CPUs. Over time the introduction of new
libraries expands our platform's addressable set of applications and use cases, unlocking new markets and increasing our long-term growth opportunity.
With ~9,000 apps, AI models, and libraries running on CUDA, NVIDIA addresses the largest market opportunity of any computing platform.
cuLitho
Computational Lithography
cuDSS
cuSPARSE
cuFFT
AmgX
Computer Aided Engineering
cuOpt
Decision Optimization
Warp
Physics
cuDF
cuML
Data Science and Processing
Megatron
Dynamo
NIXL
cuDNN
CUTLASS
Deep Learning
cuEquivariance
cuTensor
Quantum Chemistry
cuQuantum
CUDA-Q
Quantum Computing
Earth-2
Weather Analytics
MONAI
Medical Imaging
Parabricks
Genomics
Aerial
Sionna
5G/6G
Signal Processing
cuPyNumeric
Numerical Computing

**Layout**

Centered header block across the top: two-line title, then two centred paragraphs of grey body text, with about 5 percent top margin and no footer or slide number. Below, thirteen tall rounded image cards sit in a single horizontal row spanning the slide width, stepping slightly upward and shrinking from left to right, each with a centred label block (name in tan, then a grey descriptor) beneath it.

**Visuals**

Thirteen rectangular photographic or rendered cards in one row, left to right: (1) cuLitho: tall dark grey block structure with blue and gold circuit-like patterning, lit by bright blue and gold light, on a blue-black ground. (2) cuDSS: a black pipe seen in cross-section with a rainbow-coloured CFD flow stream on a blue sky background. (3) cuOpt: a purple 3D surface plot with a white line chart drawn across it, on a dark dotted grid. (4) Warp: a grey box with a white wall and coloured streamlines in red, green and blue on a dark grey floor with a grey sphere. (5) cuDF and cuML: black background with a grid and scattered coloured particle clouds in yellow, orange, cyan, pink and purple. (6) Megatron Dynamo NIXL cuDNN CUTLASS: dark teal background with a white-outlined network diagram of stacked blocks linked by lines. (7) cuEquivariance cuTensor: a ball-and-stick molecule model on black with red, blue, cyan, white and yellow atoms. (8) cuQuantum CUDA-Q: a gold dilution refrigerator cryostat with copper plates, on a dark background. (9) Earth-2: a view of Earth from space showing blue ocean, brown and green land and white clouds. (10) MONAI: colourful 3D segmented organ shapes in green, red, blue and yellow on a grey background. (11) Parabricks: a lavender double helix of DNA on a dark purple-black background. (12) Aerial Sionna: an aerial view of rooftops and streets with bright green ray paths on a grey-white city. (13) cuPyNumeric: a spiral galaxy with a purple and pink core on a deep blue starfield. Each card is a rounded-corner rectangle with a thin pale gold edge and a soft drop shadow. Below the cards is a faint, dark mirrored reflection of the row fading to black.

**Design**

Background: Solid black #000000. Pure black background (#000000) with a sans-serif NVIDIA house typeface (NVIDIA Sans). Title is white (#FFFFFF), medium weight, roughly 4.9 percent of slide height. Body paragraphs are mid grey (#A1A1A1), regular weight, roughly 2 percent of slide height. Card labels use a tan-gold (#C8B88A) medium weight for names and a light grey (#D0D0D0) regular weight for descriptors at about 1.5 percent. Cards have rounded corners of about 1 percent of slide width, thin pale gold outlines and soft dark shadows. No gradients except the fading reflection at the bottom.

**Elements and positions** (back to front; x, y, width, height as % of the slide; type, colour and alignment in brackets)

- shape at x 0, y 0, w 100, h 100 [#000000]: Full-slide black background fill (Solid fill covering the entire slide, no outline.)
- photo at x 14, y 80, w 84, h 20 [#101010]: Faint mirrored reflection of the card row fading to black along the bottom edge (Dark, low-opacity (about 15 percent) copy of the card images, fading to black; overlaps the bottom of the cards' label area.)
- photo at x 3.2, y 36, w 12.3, h 35.6 [#1E4E8C]: cuLitho card image: tall dark grey block with blue and gold circuit patterning (Rounded rectangle with thin pale gold outline and soft shadow; slight perspective tilt.)
- photo at x 15.5, y 39.2, w 10.7, h 31 [#2A6FB8]: cuDSS card image: pipe with rainbow CFD flow stream on blue sky (Rounded rectangle with thin pale gold outline and soft shadow; slight tilt.)
- photo at x 26.2, y 41.8, w 8.5, h 26 [#5A3F8A]: cuOpt card image: purple 3D surface with white line chart on dark dotted grid (Rounded rectangle with thin pale gold outline and soft shadow.)
- photo at x 34.9, y 43.5, w 7.7, h 24.6 [#4A4D52]: Warp card image: grey box with coloured streamlines on dark grey floor (Rounded rectangle with thin pale gold outline and soft shadow.)
- photo at x 42.6, y 44.8, w 7, h 22.7 [#101010]: cuDF and cuML card image: black field with coloured particle clouds (Rounded rectangle with thin pale gold outline and soft shadow.)
- photo at x 49.6, y 46, w 6.5, h 21.2 [#0E3A40]: Megatron Dynamo NIXL cuDNN CUTLASS card image: teal network diagram (Rounded rectangle with thin pale gold outline and soft shadow.)
- photo at x 56.2, y 46.6, w 6.2, h 19.9 [#000000]: cuEquivariance cuTensor card image: ball-and-stick molecule on black (Rounded rectangle with thin pale gold outline and soft shadow.)
- photo at x 62.4, y 47.2, w 6, h 18.8 [#B07A3A]: cuQuantum CUDA-Q card image: gold dilution refrigerator cryostat (Rounded rectangle with thin pale gold outline and soft shadow.)
- photo at x 68.5, y 47.6, w 5.8, h 18.5 [#2A5C9A]: Earth-2 card image: Earth from space with blue oceans and clouds (Rounded rectangle with thin pale gold outline and soft shadow.)
- photo at x 74.6, y 48.2, w 5.6, h 18 [#3A7A5A]: MONAI card image: colourful 3D segmented organs on grey (Rounded rectangle with thin pale gold outline and soft shadow.)
- photo at x 80.5, y 48.4, w 5.4, h 17.6 [#2A2233]: Parabricks card image: lavender DNA double helix on dark purple-black (Rounded rectangle with thin pale gold outline and soft shadow.)
- photo at x 85.8, y 48.7, w 5.4, h 17.5 [#8A9080]: Aerial Sionna card image: rooftops with bright green ray paths (Rounded rectangle with thin pale gold outline and soft shadow.)
- photo at x 91.2, y 48.9, w 5.3, h 17 [#1A1F4A]: cuPyNumeric card image: spiral galaxy on deep blue starfield (Rounded rectangle with thin pale gold outline and soft shadow.)
- title at x 22.7, y 5.2, w 55, h 10.2 [NVIDIA Sans, Medium, normal; font size 4.94% of slide height; #FFFFFF; center aligned]: NVIDIA CUDA-X Accelerates / Apps and Reduces the Cost of Computing (Two lines, centred, line spacing about 1.1.)
- body at x 6.9, y 18.1, w 86.6, h 4.6 [NVIDIA Sans, Regular, normal; font size 1.98% of slide height; #A1A1A1; center aligned]: CUDA-X libraries accelerate apps, enabling them to run orders of magnitude faster, more efficiently, and at lower cost than on CPUs. Over time the introduction of new / libraries expands our platform's addressable set of applications and use cases, unlocking new markets and increasing our long-term growth opportunity. (Two lines, centred, line spacing about 1.5.)
- body at x 14.2, y 24.8, w 71.6, h 2 [NVIDIA Sans, Regular, normal; font size 1.98% of slide height; #A1A1A1; center aligned]: With ~9,000 apps, AI models, and libraries running on CUDA, NVIDIA addresses the largest market opportunity of any computing platform. (One line, centred, separated from the paragraph above by a gap of about 2 percent.)
- caption at x 12, y 84, w 8, h 5 [NVIDIA Sans, Medium for name and Regular for descriptor, normal; font size 1.6% of slide height; #C8B88A; center aligned]: cuLitho / Computational Lithography (Name in tan-gold, descriptor below in light grey #D0D0D0, centred under the card.)
- caption at x 27, y 73, w 10, h 10 [NVIDIA Sans, Medium for name and Regular for descriptor, normal; font size 1.6% of slide height; #C8B88A; center aligned]: cuDSS / cuSPARSE / cuFFT / AmgX / Computer Aided Engineering (Four names stacked in tan-gold, then descriptor Computer Aided / Engineering in light grey, centred.)
- caption at x 27, y 72, w 7, h 5 [NVIDIA Sans, Medium for name and Regular for descriptor, normal; font size 1.6% of slide height; #C8B88A; center aligned]: cuOpt / Decision Optimization (Name in tan-gold, descriptor in light grey below, centred.)
- caption at x 34.5, y 71, w 8, h 5 [NVIDIA Sans, Medium for name and Regular for descriptor, normal; font size 1.6% of slide height; #C8B88A; center aligned]: Warp / Physics (Name in tan-gold, descriptor in light grey below, centred.)
- caption at x 42, y 71, w 8, h 6 [NVIDIA Sans, Medium for name and Regular for descriptor, normal; font size 1.6% of slide height; #C8B88A; center aligned]: cuDF / cuML / Data Science and Processing (Two names in tan-gold, then descriptor Data Science / and Processing in light grey, centred.)
- caption at x 49.5, y 69.5, w 8, h 11 [NVIDIA Sans, Medium for name and Regular for descriptor, normal; font size 1.6% of slide height; #C8B88A; center aligned]: Megatron / Dynamo / NIXL / cuDNN / CUTLASS / Deep Learning (Five names stacked in tan-gold, then descriptor Deep Learning in light grey, centred.)
- caption at x 62, y 69.5, w 7, h 6 [NVIDIA Sans, Medium for name and Regular for descriptor, normal; font size 1.6% of slide height; #C8B88A; center aligned]: cuEquivariance / cuTensor / Quantum Chemistry (Two names in tan-gold, then descriptor Quantum Chemistry in light grey, centred.)
- caption at x 68, y 69.5, w 6.5, h 6 [NVIDIA Sans, Medium for name and Regular for descriptor, normal; font size 1.6% of slide height; #C8B88A; center aligned]: cuQuantum / CUDA-Q / Quantum Computing (Two names in tan-gold, then descriptor Quantum Computing in light grey, centred.)
- caption at x 69.5, y 69, w 5.5, h 5 [NVIDIA Sans, Medium for name and Regular for descriptor, normal; font size 1.6% of slide height; #C8B88A; center aligned]: Earth-2 / Weather Analytics (Name in tan-gold, descriptor in light grey below, centred.)
- caption at x 75.5, y 69, w 5, h 5 [NVIDIA Sans, Medium for name and Regular for descriptor, normal; font size 1.6% of slide height; #C8B88A; center aligned]: MONAI / Medical Imaging (Name in tan-gold, descriptor in light grey below, centred.)
- caption at x 80.5, y 68.8, w 5.5, h 4.5 [NVIDIA Sans, Medium for name and Regular for descriptor, normal; font size 1.6% of slide height; #C8B88A; center aligned]: Parabricks / Genomics (Name in tan-gold, descriptor in light grey below, centred.)
- caption at x 85.5, y 69, w 5.8, h 7 [NVIDIA Sans, Medium for name and Regular for descriptor, normal; font size 1.6% of slide height; #C8B88A; center aligned]: Aerial / Sionna / 5G/6G / Signal Processing (Two names in tan-gold, then descriptor 5G/6G and Signal Processing in light grey, centred.)
- caption at x 90.8, y 68.8, w 6, h 5 [NVIDIA Sans, Medium for name and Regular for descriptor, normal; font size 1.6% of slide height; #C8B88A; center aligned]: cuPyNumeric / Numerical Computing (Name in tan-gold, descriptor Numerical Computing in light grey below, centred.)

**Measured from the file** (exact, read from the PDF rather than estimated; x, y, w, h as % of the slide; font size as % of the slide height)

- fill at x 0.0, y 0.0, w 100.0, h 100.0: #000000 (covers the slide)
- image at x 1.9, y 32.0, w 96.2, h 68.0: 6720 by 2672 pixels
- text at x 31.5, y 5.2, w 37.7, h 4.9 [NVIDIASans-Medium; font size 4.94; #FFFFFF]: NVIDIA CUDA-X Accelerates
- text at x 22.7, y 10.5, w 55.0, h 4.9 [NVIDIASans-Medium; font size 4.94; #FFFFFF]: Apps and Reduces the Cost of Computing
- text at x 6.9, y 18.1, w 86.6, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: CUDA-X libraries accelerate apps, enabling them to run orders of magnitude faster, more efficiently, and at lower cost than on CPUs. Over time the introduction of new
- text at x 10.6, y 20.7, w 79.1, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: libraries expands our platform’s addressable set of applications and use cases, unlocking new markets and increasing our long-term growth opportunity.
- text at x 14.2, y 24.8, w 71.6, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: With ~9,000 apps, AI models, and libraries running on CUDA, NVIDIA addresses the largest market opportunity of any computing platform.

### Slide 21

**Review notes**

- Review: Weaknesses are consequential: many slides depend on small gray body text and tiny labels (slides 4, 8, 13, 18, 21, 22), the financial tables on slide 9 are unreadable at presentation size, and slide 2 is a dense disclaimer wall.
- Suggested repair: Enlarge or simplify the small body text and diagram labels on slides 4, 8, 13, 18 and 21, and replace the screenshot-style tables on slide 9 with a few legible headline figures.

**Text**

NVIDIA AI Factories Are Productive, Durable, and Fungible / Productive / Profitable Tokens / Payback well inside / depreciation window / $100M / $75M / $30M / $10M / Cost / Open Models / Frontier Labs / ISP Tokens / Revenue $M/MW/year / Cost: $60M/MW build · 6-yr depreciation / Durable / CUDA-Driven Installed-Base Enhancement / Earns after full depreciation and / improves over time through software / V100 / DGX-1 / 10 yrs in service / A100 / 6 yrs old · 9 yrs sold / NVL72 / 1 yr old · 7 yrs sold / Fungible / One CUDA Architecture – Global Ecosystem / One platform, / many offtakers, any workload / AI Labs / Every Closed / and Open Model / Financial Services / Jane Street, / HRT, JPM, GS / Industrial / LG, Cadence / Pharma / Roche, Lilly / Automotive / Hyundai, BMW

**Layout**

Three equal-height outlined panels sit side by side in three columns across the lower two-thirds of the slide, each with a green column heading and white subheading above it. The title is centred across the top at about 10.5% from the top edge, and the NVIDIA logo sits bottom right. Left and right margins are about 5%, and the panels are separated by gaps of about 1.5%.

**Visuals**

Three panels. Left panel: vertical bar chart with title 'Payback well inside depreciation window' and a y-axis that is not drawn. Bars from left to right: Cost $10M (grey #757575), Open Models $30M (dark green about #3E6600), Frontier Labs $75M (green #5E9000, #76B900 at 80 percent opacity), ISP Tokens $100M (lime #B8FF3C). Values are printed above each bar. Bars sit on a thin grey baseline at about 76.5% of slide height; axis labels are printed under each bar. Footnote under the chart gives the units and assumptions. The chart has no legend and no gridlines. ISP Tokens is the highlighted bar. Middle panel: three horizontal stacked bars on a shared 0 to full-width scale with no axis. V100 / DGX-1: one full-width bar in #76B900 labelled 10 yrs in service. A100: green segment from about 38.1% to 52.4% of slide width, dark green #3E6600 segment to about 59.5%, grey #757575 segment to about 62.0%, labelled 6 yrs old · 9 yrs sold. NVL72: green segment 38.1% to about 40.5%, dark green to about 54.8%, grey to about 62.0%, labelled 1 yr old · 7 yrs sold. Each bar has a dark track #2A2E31 behind it. Right panel: six line icons in lime green #76B900 with bold white labels. AI Labs icon shows three people figures above a stacked group. Financial Services icon shows a rising line chart over three bars with a dollar coin. Industrial icon shows a factory with chimneys. Pharma icon shows a medicine bottle with a cross and a capsule. Automotive icon shows a side-view car, with a thin green leader line crossing it from the Financial Services label area. All icons are outline style with about 0.2% stroke weight.

**Design**

Background: Solid black #000000 covering the whole slide. Pure black background. Title in white #FFFFFF, subtitles in white #FFFFFF, column headings in NVIDIA green #76B900, and body and labels in white #FFFFFF or grey #989898 and #A1A1A1. Accents are #76B900, dark green #3E6600, lime #B8FF3C and grey #757575. Typeface is NVIDIASans, a clean geometric sans-serif, in Medium for the title and headings, Regular for body text and Bold for category labels. Title is about 4.9% of slide height, column headings about 2.5%, subheads and panel headlines about 2.2%, and body and labels about 1.5% to 2.0%. Panels have 1px grey #5A5A5A rounded outlines with corner radii of about 0.5% of slide height and no fill or shadow.

**Elements and positions** (back to front; x, y, width, height as % of the slide; type, colour and alignment in brackets)

- shape at x 0, y 0, w 100, h 100: Full-slide background rectangle, solid black (fill #000000 covering the slide)
- shape at x 5, y 31, w 29, h 55: Left panel outline, rounded rectangle (no fill, 1px stroke #5A5A5A, corner radius about 0.5% of slide height)
- shape at x 35.5, y 31, w 29, h 55: Middle panel outline, rounded rectangle (no fill, 1px stroke #5A5A5A, corner radius about 0.5% of slide height)
- shape at x 66, y 31, w 29, h 55: Right panel outline, rounded rectangle (no fill, 1px stroke #5A5A5A, corner radius about 0.5% of slide height)
- shape at x 8.5, y 73.6, w 2.8, h 2.9: Cost bar, rectangle (fill #757575, no outline)
- shape at x 14.6, y 67.6, w 2.7, h 8.9: Open Models bar, rectangle (fill about #3E6600, no outline)
- shape at x 20.7, y 54.2, w 2.7, h 22.4 [#76B900]: Frontier Labs bar, rectangle (fill #76B900 at 80 percent opacity)
- shape at x 26.7, y 46.7, w 2.7, h 29.9 [#B8FF3C]: ISP Tokens bar, rectangle (fill #B8FF3C, highlighted bar, no outline)
- line at x 6.9, y 76.4, w 24.2, h 0.2: Chart baseline, horizontal line (stroke grey #757575, about 1px)
- shape at x 38.1, y 49.4, w 23.8, h 2.6 [#2A2E31]: V100 / DGX-1 track, rectangle (fill #2A2E31, no outline)
- shape at x 38.1, y 49.4, w 23.8, h 2.6 [#76B900]: V100 / DGX-1 bar, rectangle (fill #76B900, full width, no outline)
- shape at x 38.1, y 62.2, w 23.8, h 2.6 [#757575]: A100 grey track, rectangle (fill #757575, no outline)
- shape at x 38.1, y 62.2, w 21.4, h 2.6 [#3E6600]: A100 dark green segment, rectangle (fill #3E6600, no outline)
- shape at x 38.1, y 62.2, w 14.2, h 2.6 [#76B900]: A100 green segment, rectangle (fill #76B900, no outline)
- shape at x 38.1, y 75, w 23.8, h 2.6 [#757575]: NVL72 grey track, rectangle (fill #757575, no outline)
- shape at x 38.1, y 75, w 16.7, h 2.6 [#3E6600]: NVL72 dark green segment, rectangle (fill #3E6600, no outline)
- shape at x 38.1, y 75, w 2.4, h 2.6 [#76B900]: NVL72 green segment, rectangle (fill #76B900, no outline)
- icon at x 72.5, y 47.2, w 4.7, h 5: AI Labs icon: three people figures above a grouped base (lime #76B900 outline line art, about 0.2% stroke)
- icon at x 83.8, y 46.4, w 3.3, h 5.8: Financial Services icon: rising line over three bars with dollar coin (lime #76B900 outline line art, about 0.2% stroke)
- icon at x 69.1, y 65.2, w 4.9, h 5.7: Industrial icon: factory with chimneys (lime #76B900 outline line art, about 0.2% stroke)
- icon at x 78.6, y 65.2, w 3.5, h 6.3: Pharma icon: medicine bottle with cross and capsule (lime #76B900 outline line art, about 0.2% stroke)
- icon at x 86.9, y 65.8, w 5.2, h 4.4: Automotive icon: side-view car (lime #76B900 outline line art, about 0.2% stroke, crossed by leader line)
- line at x 88.1, y 61.3, w 1.3, h 8.2 [#76B900]: Leader line crossing the Automotive icon (thin lime #76B900 stroke about 0.2% width, angled from upper left to lower right)
- logo at x 92.9, y 94, w 4.6, h 2 [#76B900]: NVIDIA logo, bottom right corner, green eye mark with NVIDIA wordmark (eye mark in #76B900, wordmark in white #FFFFFF)
- title at x 12.3, y 10.5, w 75.3, h 4.9 [NVIDIASans Medium, sans-serif, regular case; font size 4.94% of slide height; #FFFFFF; center aligned]: NVIDIA AI Factories Are Productive, Durable, and Fungible
- subtitle at x 15.9, y 24.3, w 7.1, h 2.5 [NVIDIASans Medium, sans-serif; font size 2.47% of slide height; #76B900; center aligned]: Productive
- subtitle at x 47.4, y 24.3, w 5.1, h 2.5 [NVIDIASans Medium, sans-serif; font size 2.47% of slide height; #76B900; center aligned]: Durable
- subtitle at x 77.8, y 24.3, w 5.6, h 2.5 [NVIDIASans Medium, sans-serif; font size 2.47% of slide height; #76B900; center aligned]: Fungible
- body at x 14.8, y 27.4, w 9.2, h 2 [NVIDIASans Regular, sans-serif; font size 1.98% of slide height; #FFFFFF; center aligned]: Profitable Tokens
- body at x 39, y 27.4, w 22, h 2 [NVIDIASans Regular, sans-serif; font size 1.98% of slide height; #FFFFFF; center aligned]: CUDA-Driven Installed-Base Enhancement
- body at x 69.3, y 27.4, w 22.9, h 2 [NVIDIASans Regular, sans-serif; font size 1.98% of slide height; #FFFFFF; center aligned]: One CUDA Architecture – Global Ecosystem
- body at x 13.3, y 35.6, w 12.3, h 4.6 [NVIDIASans Medium, sans-serif; font size 2.22% of slide height; #FFFFFF; center aligned]: Payback well inside / depreciation window
- body at x 39, y 35.6, w 22.1, h 4.6 [NVIDIASans Medium, sans-serif; font size 2.22% of slide height; #FFFFFF; center aligned]: Earns after full depreciation and / improves over time through software
- body at x 71.7, y 35.6, w 17.6, h 4.6 [NVIDIASans Medium, sans-serif; font size 2.22% of slide height; #FFFFFF; center aligned]: One platform, / many offtakers, any workload
- label at x 26.7, y 44.8, w 2.8, h 1.5 [NVIDIASans Regular, sans-serif; font size 1.48% of slide height; #FFFFFF; center aligned]: $100M
- label at x 20.9, y 52.3, w 2.3, h 1.5 [NVIDIASans Regular, sans-serif; font size 1.48% of slide height; #FFFFFF; center aligned]: $75M
- label at x 14.8, y 65.8, w 2.3, h 1.5 [NVIDIASans Regular, sans-serif; font size 1.48% of slide height; #FFFFFF; center aligned]: $30M
- label at x 8.8, y 71.7, w 2.3, h 1.5 [NVIDIASans Regular, sans-serif; font size 1.48% of slide height; #FFFFFF; center aligned]: $10M
- label at x 9, y 77.8, w 1.8, h 1.5 [NVIDIASans Regular, sans-serif; font size 1.48% of slide height; #FFFFFF; center aligned]: Cost
- label at x 13.4, y 77.8, w 11.2, h 1.5 [NVIDIASans Regular, sans-serif; font size 1.48% of slide height; #FFFFFF; center aligned]: Open Models Frontier Labs (two category labels, Open Models under the second bar and Frontier Labs under the third bar)
- label at x 25.9, y 77.8, w 4.3, h 1.5 [NVIDIASans Regular, sans-serif; font size 1.48% of slide height; #FFFFFF; center aligned]: ISP Tokens
- caption at x 12.9, y 81.1, w 13.1, h 2.4 [NVIDIASans Regular, sans-serif; font size 1.24% of slide height; #757575; center aligned]: Revenue $M/MW/year / Cost: $60M/MW build · 6-yr depreciation
- label at x 38.1, y 46.3, w 8.1, h 2.2 [NVIDIASans Medium, sans-serif; font size 2.22% of slide height; #FFFFFF; left aligned]: V100 / DGX-1
- label at x 54.6, y 46.8, w 7.3, h 1.7 [NVIDIASans Regular, sans-serif; font size 1.73% of slide height; #76B900; right aligned]: 10 yrs in service
- label at x 38.1, y 59, w 3.1, h 2.2 [NVIDIASans Medium, sans-serif; font size 2.22% of slide height; #FFFFFF; left aligned]: A100
- label at x 53, y 59.6, w 8.9, h 1.7 [NVIDIASans Regular, sans-serif; font size 1.73% of slide height; #76B900; right aligned]: 6 yrs old · 9 yrs sold
- label at x 38.1, y 71.8, w 3.9, h 2.2 [NVIDIASans Medium, sans-serif; font size 2.22% of slide height; #FFFFFF; left aligned]: NVL72
- label at x 53.5, y 72.4, w 8.4, h 1.7 [NVIDIASans Regular, sans-serif; font size 1.73% of slide height; #76B900; right aligned]: 1 yr old · 7 yrs sold
- label at x 73.3, y 53.6, w 3.9, h 2 [NVIDIASans Bold, sans-serif; font size 1.98% of slide height; #FFFFFF; center aligned]: AI Labs
- label at x 81, y 53.6, w 9.5, h 2 [NVIDIASans Bold, sans-serif; font size 1.98% of slide height; #FFFFFF; center aligned]: Financial Services
- caption at x 71.5, y 56.1, w 8, h 3.7 [NVIDIASans Regular, sans-serif; font size 1.73% of slide height; #989898; center aligned]: Every Closed / and Open Model
- caption at x 82.7, y 56.1, w 6.1, h 3.7 [NVIDIASans Regular, sans-serif; font size 1.73% of slide height; #989898; center aligned]: Jane Street, / HRT, JPM, GS
- label at x 68.7, y 73, w 5.1, h 2 [NVIDIASans Bold, sans-serif; font size 1.98% of slide height; #FFFFFF; center aligned]: Industrial
- label at x 78.3, y 73, w 4.1, h 2 [NVIDIASans Bold, sans-serif; font size 1.98% of slide height; #FFFFFF; center aligned]: Pharma
- label at x 86.4, y 73, w 6.2, h 2 [NVIDIASans Bold, sans-serif; font size 1.98% of slide height; #FFFFFF; center aligned]: Automotive
- caption at x 68.5, y 75.5, w 5.6, h 1.7 [NVIDIASans Regular, sans-serif; font size 1.73% of slide height; #989898; center aligned]: LG, Cadence
- caption at x 77.8, y 75.3, w 5, h 1.7 [NVIDIASans Regular, sans-serif; font size 1.73% of slide height; #989898; center aligned]: Roche, Lilly
- caption at x 86.2, y 75.5, w 6.7, h 1.7 [NVIDIASans Regular, sans-serif; font size 1.73% of slide height; #989898; center aligned]: Hyundai, BMW

**Measured from the file** (exact, read from the PDF rather than estimated; x, y, w, h as % of the slide; font size as % of the slide height)

- fill at x 0.0, y 0.0, w 100.0, h 100.0: #000000 (covers the slide)
- fill at x 20.7, y 54.2, w 2.7, h 22.4: #76B900, opacity 0.80
- fill at x 26.7, y 46.7, w 2.7, h 29.9: #B8FF3C
- fill at x 38.1, y 49.4, w 23.8, h 2.6: #2A2E31
- fill at x 38.1, y 49.4, w 23.8, h 2.6: #76B900
- fill at x 38.1, y 62.2, w 23.8, h 2.6: #757575
- fill at x 38.1, y 62.2, w 21.4, h 2.6: #3E6600
- fill at x 38.1, y 75.0, w 23.8, h 2.6: #757575
- fill at x 38.1, y 75.0, w 16.7, h 2.6: #3E6600
- text at x 12.3, y 10.5, w 75.3, h 4.9 [NVIDIASans-Medium; font size 4.94; #FFFFFF]: NVIDIA AI Factories Are Productive, Durable, and Fungible
- text at x 15.9, y 24.3, w 7.1, h 2.5 [NVIDIASans-Medium; font size 2.47; #76B900]: Productive
- text at x 47.4, y 24.3, w 5.1, h 2.5 [NVIDIASans-Medium; font size 2.47; #76B900]: Durable
- text at x 77.8, y 24.3, w 5.6, h 2.5 [NVIDIASans-Medium; font size 2.47; #76B900]: Fungible
- text at x 14.8, y 27.4, w 9.2, h 2.0 [NVIDIASans-Regular; font size 1.98; #FFFFFF]: Profitable Tokens
- text at x 39.0, y 27.4, w 22.0, h 2.0 [NVIDIASans-Regular; font size 1.98; #FFFFFF]: CUDA-Driven Installed-Base Enhancement
- text at x 69.3, y 27.4, w 22.9, h 2.0 [NVIDIASans-Regular; font size 1.98; #FFFFFF]: One CUDA Architecture – Global Ecosystem
- text at x 13.7, y 35.6, w 11.4, h 2.2 [NVIDIASans-Medium; font size 2.22; #FFFFFF]: Payback well inside
- text at x 40.5, y 35.6, w 19.4, h 2.2 [NVIDIASans-Medium; font size 2.22; #FFFFFF]: Earns after full depreciation and
- text at x 76.4, y 35.6, w 8.5, h 2.2 [NVIDIASans-Medium; font size 2.22; #FFFFFF]: One platform,
- text at x 13.3, y 38.0, w 12.3, h 2.2 [NVIDIASans-Medium; font size 2.22; #FFFFFF]: depreciation window
- text at x 39.0, y 38.0, w 22.1, h 2.2 [NVIDIASans-Medium; font size 2.22; #FFFFFF]: improves over time through software
- text at x 71.7, y 38.0, w 17.6, h 2.2 [NVIDIASans-Medium; font size 2.22; #FFFFFF]: many offtakers, any workload
- text at x 38.1, y 46.3, w 8.1, h 2.2 [NVIDIASans-Medium; font size 2.22; #FFFFFF]: V100 / DGX-1
- text at x 38.1, y 59.0, w 3.1, h 2.2 [NVIDIASans-Medium; font size 2.22; #FFFFFF]: A100
- text at x 54.6, y 46.8, w 7.3, h 1.7 [NVIDIASans-Regular; font size 1.73; #76B900]: 10 yrs in service
- text at x 73.3, y 53.6, w 3.9, h 2.0 [NVIDIASans-Bold; font size 1.98; #FFFFFF]: AI Labs
- text at x 81.0, y 53.6, w 9.5, h 2.0 [NVIDIASans-Bold; font size 1.98; #FFFFFF]: Financial Services
- text at x 26.7, y 44.8, w 2.8, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: $100M
- text at x 38.1, y 71.8, w 3.9, h 2.2 [NVIDIASans-Medium; font size 2.22; #FFFFFF]: NVL72
- text at x 72.3, y 56.1, w 6.0, h 1.7 [NVIDIASans-Regular; font size 1.73; #989898]: Every Closed
- text at x 83.0, y 56.1, w 5.8, h 1.7 [NVIDIASans-Regular; font size 1.73; #989898]: Jane Street,
- text at x 71.5, y 58.1, w 4.6, h 1.7 [NVIDIASans-Regular; font size 1.73; #989898]: and Open
- text at x 76.1, y 58.1, w 2.8, h 1.7 [NVIDIASans-Regular; font size 1.73; #A1A1A1]: Model
- text at x 82.7, y 58.1, w 6.1, h 1.7 [NVIDIASans-Regular; font size 1.73; #989898]: HRT, JPM, GS
- text at x 53.0, y 59.6, w 8.9, h 1.7 [NVIDIASans-Regular; font size 1.73; #76B900]: 6 yrs old · 9 yrs sold
- text at x 20.9, y 52.3, w 2.3, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: $75M
- text at x 68.7, y 73.0, w 5.1, h 2.0 [NVIDIASans-Bold; font size 1.98; #FFFFFF]: Industrial
- text at x 78.3, y 73.0, w 4.1, h 2.0 [NVIDIASans-Bold; font size 1.98; #FFFFFF]: Pharma
- text at x 86.4, y 73.0, w 6.2, h 2.0 [NVIDIASans-Bold; font size 1.98; #FFFFFF]: Automotive
- text at x 53.5, y 72.4, w 8.4, h 1.7 [NVIDIASans-Regular; font size 1.73; #76B900]: 1 yr old · 7 yrs sold
- text at x 68.5, y 75.5, w 5.6, h 1.7 [NVIDIASans-Regular; font size 1.73; #989898]: LG, Cadence
- text at x 77.8, y 75.3, w 5.0, h 1.7 [NVIDIASans-Regular; font size 1.73; #989898]: Roche, Lilly
- text at x 86.2, y 75.5, w 6.7, h 1.7 [NVIDIASans-Regular; font size 1.73; #989898]: Hyundai, BMW
- text at x 14.8, y 65.8, w 2.3, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: $30M
- text at x 8.8, y 71.7, w 2.3, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: $10M
- text at x 9.0, y 77.8, w 1.8, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Cost
- text at x 13.4, y 77.8, w 11.2, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Open Models Frontier Labs
- text at x 25.9, y 77.8, w 4.3, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: ISP Tokens
- text at x 15.8, y 81.1, w 7.2, h 1.2 [NVIDIASans-Regular; font size 1.24; #757575]: Revenue $M/MW/year
- text at x 12.9, y 82.4, w 13.1, h 1.2 [NVIDIASans-Regular; font size 1.24; #757575]: Cost: $60M/MW build · 6-yr depreciation

### Slide 22

**Review notes**

- Worth studying: A residual value curve by GPU generation against an accelerated depreciation line makes an accounting argument visual and easy to follow.
- Review: Weaknesses are consequential: many slides depend on small gray body text and tiny labels (slides 4, 8, 13, 18, 21, 22), the financial tables on slide 9 are unreadable at presentation size, and slide 2 is a dense disclaimer wall.

**Text**

NVIDIA AI Infrastructure Retains Value
Beyond Accelerated Depreciation Schedules
8-GPU HGX System Cost Per GPU · Dotted Curve = Accelerated Schedule, 25% Residual at Year 2, Zero by Year 5
Residual Value, % of Purchase Price
B200 158%
150%
100%
Accelerated Depreciation
H100 58%
50%
25% at Year 2
A100 25%
0%
0
1
2
3
4
5
6
7
8
Years Since Mass Availability
Residual value: Silicon Data, monthly averages. Purchase price basis: A100 $20,000 · H100 $35,000 · B200 $45,000.
Mass-availability dates assumed; purchase prices are third-party estimates, not Silicon Data data.

**Layout**

Centred stack on a black slide: a two-line centred title at the top, a grey centred subtitle beneath it, then one large rounded-outline chart panel spanning roughly 13% to 87% of the width and from about 23% to 90% of the height. Two-line source footer sits bottom left at about 1.7% margin and the NVIDIA logo sits bottom right; no slide number is visible.

**Visuals**

One line chart titled 'Residual Value, % of Purchase Price', with the x axis 'Years Since Mass Availability' running 0 to 8 (ticks every 1 year, unit years) and the y axis 'Residual Value, % of Purchase Price' running 0% to about 160% (labelled ticks 0%, 50%, 100%, 150%, no unit beyond %). Series: (1) Dotted white curve 'Accelerated Depreciation' from about (0 yr, 100%) falling convexly to (2 yr, 25%, labelled '25% at Year 2' with a small white dot marker) and reaching (5 yr, 0%); it is a decay curve, values at other points about: 1 yr about 50%, 3 yr about 12%, 4 yr about 6%. (2) Orange solid wavy line 'B200 158%' from about 0.8 yr at about 138% rising to a peak about 1.4 yr near 162%, ending at about 1.5 yr at 158%, labelled at its end. (3) Blue solid wavy line 'H100 58%' from about 2.5 yr at about 55% dipping to about 43% at 3.2 yr, rising to a peak about 62% at 3.7 yr and ending at 4.0 yr at 58%, labelled at its end. (4) Green solid wavy line 'A100 25%' from about 4.8 yr at about 41% declining in steps to about 27% near 6 yr and ending at about 6.3 yr at 25%, labelled at its end. No legend; series are identified by end labels. Highlighted point: the white dot at 2 yr, 25% on the dotted curve. Faint horizontal gridlines at 50%, 100% and 150%, with a slightly brighter baseline at 0%.

**Design**

Background: Solid black #000000. Pure black background, no gradients or textures. Title white #FFFFFF in NVIDIASans Medium, a clean geometric sans-serif, about 4.9% of slide height per line, centred. Subtitle #A1A1A1 in NVIDIASans Regular, about 2.0% of height. Chart title and axis labels white #FFFFFF or #EDEDED in NVIDIASans Regular, about 1.7% to 2.0% of height. Footer #757575 in NVIDIASans Regular, about 1.2% of height. Accent colours: orange #F5A623 for B200, blue #4DA3FF for H100, green #76B900 for A100 and the NVIDIA logo, white dotted line for the accelerated schedule. Chart panel is a thin 1px #8C8C8C rounded rectangle outline with a small corner radius of about 0.6% of slide width, no fill and no shadow. Gridlines are hairlines about #1F1F1F.

**Elements and positions** (back to front; x, y, width, height as % of the slide; type, colour and alignment in brackets)

- shape at x 0, y 0, w 100, h 100 [#000000]: Full-slide background rectangle (Solid fill, no outline, covers the whole slide.)
- shape at x 12.9, y 22.9, w 74.2, h 67.3 [#8C8C8C]: Chart panel outline, rounded rectangle (No fill (transparent), 1px stroke #8C8C8C, corner radius about 0.6% of slide width, no shadow; encloses the chart title, plot and axis label.)
- text at x 40.8, y 25.8, w 18.3, h 2 [NVIDIASans Regular, normal weight; font size 1.98% of slide height; #FFFFFF; center aligned]: Residual Value, % of Purchase Price
- line at x 19.5, y 35.6, w 64.3, h 29.9 [#1F1F1F]: Horizontal gridlines at 150%, 100%, 50% (Three hairline gridlines about 0.1% of height thick at y of 150%, 100% and 50% ticks; no fill.)
- line at x 19.5, y 80.4, w 64.3, h 0.1 [#3A3A3A]: Baseline at 0% (Hairline across the plot bottom, slightly brighter than gridlines.)
- chart at x 16, y 31.5, w 68.5, h 51 [#FFFFFF]: Line chart plot area: Residual Value % versus Years Since Mass Availability, four series (Contains the dotted accelerated curve, the B200, H100 and A100 curves, end labels, y ticks 0% to 150% and x ticks 0 to 8.)
- line at x 19.5, y 50.7, w 40.1, h 29.8 [#FFFFFF]: Dotted accelerated depreciation curve, 100% at 0 yr to 0% at 5 yr (Dotted white stroke about 1px, dot pitch about 2px, convex decay curve; passes through a white marker dot at 2 yr, 25%.)
- shape at x 35.4, y 72.7, w 0.4, h 0.6 [#FFFFFF]: Marker dot at 2 yr, 25% (Small filled white circle about 4px diameter on the dotted curve.)
- line at x 25.9, y 31.6, w 5.7, h 7.7 [#F5A623]: B200 series curve (orange) (Solid smooth wavy stroke about 2.5px wide, from about 0.8 yr at 138% to 1.5 yr at 158%.)
- line at x 39.5, y 61.3, w 12.2, h 6.4 [#4DA3FF]: H100 series curve (blue) (Solid smooth wavy stroke about 2.5px wide, from 2.5 yr at about 55% to 4.0 yr at 58%.)
- line at x 58, y 68, w 12.3, h 5.3 [#76B900]: A100 series curve (green) (Solid stepped wavy stroke about 2.5px wide, from 4.8 yr at about 41% to 6.3 yr at 25%.)
- text at x 31.9, y 32.5, w 5.4, h 1.7 [NVIDIASans Regular, normal weight; font size 1.73% of slide height; #EDEDED; left aligned]: B200 158%
- text at x 16, y 34.8, w 2.5, h 1.7 [NVIDIASans Regular, normal weight; font size 1.73% of slide height; #EDEDED; right aligned]: 150%
- text at x 16, y 49.7, w 2.5, h 1.7 [NVIDIASans Regular, normal weight; font size 1.73% of slide height; #EDEDED; right aligned]: 100%
- text at x 24.8, y 58.6, w 11.4, h 1.7 [NVIDIASans Regular, normal weight; font size 1.73% of slide height; #EDEDED; left aligned]: Accelerated Depreciation
- text at x 51.9, y 62.4, w 4.9, h 1.7 [NVIDIASans Regular, normal weight; font size 1.73% of slide height; #EDEDED; left aligned]: H100 58%
- text at x 16.6, y 64.7, w 2, h 1.7 [NVIDIASans Regular, normal weight; font size 1.73% of slide height; #EDEDED; right aligned]: 50%
- text at x 34.4, y 69.9, w 6.2, h 1.7 [NVIDIASans Regular, normal weight; font size 1.73% of slide height; #EDEDED; left aligned]: 25% at Year 2
- text at x 70.4, y 72.3, w 4.9, h 1.7 [NVIDIASans Regular, normal weight; font size 1.73% of slide height; #EDEDED; left aligned]: A100 25%
- text at x 17.2, y 79.6, w 1.4, h 1.7 [NVIDIASans Regular, normal weight; font size 1.73% of slide height; #EDEDED; right aligned]: 0%
- text at x 19.2, y 82, w 0.6, h 1.7 [NVIDIASans Regular, normal weight; font size 1.73% of slide height; #EDEDED; center aligned]: 0
- text at x 27.3, y 82, w 0.6, h 1.7 [NVIDIASans Regular, normal weight; font size 1.73% of slide height; #EDEDED; center aligned]: 1
- text at x 35.3, y 82, w 0.6, h 1.7 [NVIDIASans Regular, normal weight; font size 1.73% of slide height; #EDEDED; center aligned]: 2
- text at x 43.3, y 82, w 0.6, h 1.7 [NVIDIASans Regular, normal weight; font size 1.73% of slide height; #EDEDED; center aligned]: 3
- text at x 51.3, y 82, w 0.6, h 1.7 [NVIDIASans Regular, normal weight; font size 1.73% of slide height; #EDEDED; center aligned]: 4
- text at x 59.4, y 82, w 0.6, h 1.7 [NVIDIASans Regular, normal weight; font size 1.73% of slide height; #EDEDED; center aligned]: 5
- text at x 67.4, y 82, w 0.6, h 1.7 [NVIDIASans Regular, normal weight; font size 1.73% of slide height; #EDEDED; center aligned]: 6
- text at x 75.4, y 82, w 0.6, h 1.7 [NVIDIASans Regular, normal weight; font size 1.73% of slide height; #EDEDED; center aligned]: 7
- text at x 83.5, y 82, w 0.6, h 1.7 [NVIDIASans Regular, normal weight; font size 1.73% of slide height; #EDEDED; center aligned]: 8
- text at x 43.5, y 86.7, w 13, h 1.7 [NVIDIASans Regular, normal weight; font size 1.73% of slide height; #FFFFFF; center aligned]: Years Since Mass Availability
- text at x 21.3, y 5.2, w 57.8, h 10.2 [NVIDIASans Medium, normal weight, sans-serif, no capitals transform, default letter-spacing; font size 4.94% of slide height; #FFFFFF; center aligned]: NVIDIA AI Infrastructure Retains Value / Beyond Accelerated Depreciation Schedules (Two lines, line pitch about 5.3% of height.)
- text at x 21.2, y 18.1, w 57.7, h 2 [NVIDIASans Regular, normal weight, sans-serif; font size 1.98% of slide height; #A1A1A1; center aligned]: 8-GPU HGX System Cost Per GPU · Dotted Curve = Accelerated Schedule, 25% Residual at Year 2, Zero by Year 5
- text at x 1.7, y 94.8, w 36.9, h 2.5 [NVIDIASans Regular, normal weight, sans-serif; font size 1.24% of slide height; #757575; left aligned]: Residual value: Silicon Data, monthly averages. Purchase price basis: A100 $20,000 · H100 $35,000 · B200 $45,000. / Mass-availability dates assumed; purchase prices are third-party estimates, not Silicon Data data. (Two lines at about 1.2% of height, line pitch about 1.3% of height.)
- logo at x 92.9, y 94, w 4.6, h 1.8 [#76B900; right aligned]: NVIDIA logo: green eye mark with white NVIDIA wordmark (Eye mark in #76B900 green, wordmark in white #FFFFFF, no background, bottom right corner.)

**Measured from the file** (exact, read from the PDF rather than estimated; x, y, w, h as % of the slide; font size as % of the slide height)

- fill at x 0.0, y 0.0, w 100.0, h 100.0: #000000 (covers the slide)
- text at x 24.7, y 5.2, w 51.3, h 4.9 [NVIDIASans-Medium; font size 4.94; #FFFFFF]: NVIDIA AI Infrastructure Retains Value
- text at x 21.3, y 10.5, w 57.8, h 4.9 [NVIDIASans-Medium; font size 4.94; #FFFFFF]: Beyond Accelerated Depreciation Schedules
- text at x 21.2, y 18.1, w 57.7, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: 8-GPU HGX System Cost Per GPU · Dotted Curve = Accelerated Schedule, 25% Residual at Year 2, Zero by Year 5
- text at x 40.8, y 25.8, w 18.3, h 2.0 [NVIDIASans-Regular; font size 1.98; #FFFFFF]: Residual Value, % of Purchase Price
- text at x 31.9, y 32.5, w 5.4, h 1.7 [NVIDIASans-Regular; font size 1.73; #EDEDED]: B200 158%
- text at x 16.0, y 34.8, w 2.5, h 1.7 [NVIDIASans-Regular; font size 1.73; #EDEDED]: 150%
- text at x 16.0, y 49.7, w 2.5, h 1.7 [NVIDIASans-Regular; font size 1.73; #EDEDED]: 100%
- text at x 24.8, y 58.6, w 11.4, h 1.7 [NVIDIASans-Regular; font size 1.73; #EDEDED]: Accelerated Depreciation
- text at x 51.9, y 62.4, w 4.9, h 1.7 [NVIDIASans-Regular; font size 1.73; #EDEDED]: H100 58%
- text at x 16.6, y 64.7, w 2.0, h 1.7 [NVIDIASans-Regular; font size 1.73; #EDEDED]: 50%
- text at x 34.4, y 69.9, w 6.2, h 1.7 [NVIDIASans-Regular; font size 1.73; #EDEDED]: 25% at Year 2
- text at x 70.4, y 72.3, w 4.9, h 1.7 [NVIDIASans-Regular; font size 1.73; #EDEDED]: A100 25%
- text at x 17.2, y 79.6, w 1.4, h 1.7 [NVIDIASans-Regular; font size 1.73; #EDEDED]: 0%
- text at x 19.2, y 82.0, w 0.6, h 1.7 [NVIDIASans-Regular; font size 1.73; #EDEDED]: 0
- text at x 27.3, y 82.0, w 0.6, h 1.7 [NVIDIASans-Regular; font size 1.73; #EDEDED]: 1
- text at x 35.3, y 82.0, w 0.6, h 1.7 [NVIDIASans-Regular; font size 1.73; #EDEDED]: 2
- text at x 43.3, y 82.0, w 0.6, h 1.7 [NVIDIASans-Regular; font size 1.73; #EDEDED]: 3
- text at x 51.3, y 82.0, w 0.6, h 1.7 [NVIDIASans-Regular; font size 1.73; #EDEDED]: 4
- text at x 59.4, y 82.0, w 0.6, h 1.7 [NVIDIASans-Regular; font size 1.73; #EDEDED]: 5
- text at x 67.4, y 82.0, w 0.6, h 1.7 [NVIDIASans-Regular; font size 1.73; #EDEDED]: 6
- text at x 75.4, y 82.0, w 0.6, h 1.7 [NVIDIASans-Regular; font size 1.73; #EDEDED]: 7
- text at x 83.5, y 82.0, w 0.6, h 1.7 [NVIDIASans-Regular; font size 1.73; #EDEDED]: 8
- text at x 43.5, y 86.7, w 13.0, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: Years Since Mass Availability
- text at x 1.7, y 94.8, w 36.9, h 1.2 [NVIDIASans-Regular; font size 1.24; #757575]: Residual value: Silicon Data, monthly averages. Purchase price basis: A100 $20,000 · H100 $35,000 · B200 $45,000.
- text at x 1.7, y 96.1, w 31.5, h 1.2 [NVIDIASans-Regular; font size 1.24; #757575]: Mass-availability dates assumed; purchase prices are third-party estimates, not Silicon Data data.

### Slide 23

**Text**

Open Models Enable New Builders and Create New Markets
Advancements in open models are contributing to the usefulness and proliferation of AI.
Open-weight models run primarily on NVIDIA and account for ~75% of token generation today, up from ~40% a year ago.
Token consumption is growing exponentially – up 25X over the past year.
Open Router Token Volume - Open vs. Closed
Proprietary
Open-Weight
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
Jan-25
Feb-25
Mar-25
Apr-25
May-25
Jun-25
Jul-25
Aug-25
Sep-25
Oct-25
Nov-25
Dec-25
Jan-26
Feb-26
Mar-26
Apr-26
May-26
Jun-26
Jul-26
Aug-26
Open Router Monthly Token Traffic
(Trillions)
400
350
300
250
200
150
100
50
0
2
3
6
8
10
9
11
15
21
22
27
26
32
50
84
97
123
188
248
379
Jan
2025
Apr
2025
Jul
2025
Oct
2025
Jan
2026
Apr
2026
Jul
2026
Source: OpenRouter
NVIDIA

**Layout**

Single centred title across the top at about 10.5% from the top, with a three-line centred subtitle block beneath it. Below, two equal-width rounded panels sit side by side on a black field, the left panel holding a 100% stacked area chart and the right panel a line chart with a 'Source: OpenRouter' line at its lower right; the NVIDIA logo sits bottom right, with no slide number or footer.

**Visuals**

Two charts. (1) Left chart: 100% stacked area chart titled 'Open Router Token Volume - Open vs. Closed', legend 'Proprietary' (blue square #0074DF) and 'Open-Weight' (green square #76B900). Y axis 0% to 100% in 10% steps, no unit beyond percent (share of token volume). X axis monthly categories Jan-25 to Aug-26, labels rotated about 45 degrees. Proprietary series (bottom, blue) with about values: Jan-25 ~76%, Feb-25 ~79%, Mar-25 ~76%, Apr-25 ~75%, May-25 ~80%, Jun-25 ~77%, Jul-25 ~69%, Aug-25 ~62%, Sep-25 ~76%, Oct-25 ~76%, Nov-25 ~78%, Dec-25 ~70%, Jan-26 ~67%, Feb-26 ~50%, Mar-26 ~57%, Apr-26 ~61%, May-26 ~42%, Jun-26 ~32%, Jul-26 ~24%, Aug-26 ~27%. Open-Weight series (top, green) fills the remainder to 100%; no values are printed. No highlighted points. (2) Right chart: line chart titled 'Open Router Monthly Token Traffic' with subtitle '(Trillions)'. Y axis 0 to 400 in steps of 50, unit trillions (tokens). X axis labelled Jan 2025, Apr 2025, Jul 2025, Oct 2025, Jan 2026, Apr 2026, Jul 2026 at quarterly intervals, baseline grey. One green line (#76B900, about 2.5pt, no markers) with printed value labels in white above each point, in order: 2, 3, 6, 8, 10, 9, 11, 15, 21, 22, 27, 26, 32, 50, 84, 97, 123, 188, 248, 379. Rises from about 2 at the left to 379 at the right end, the steepest rise in the final months. No legend; no highlighted points. Source line 'Source: OpenRouter' in grey at lower right of the panel. (3) NVIDIA logo (green eye mark with white NVIDIA wordmark) bottom right.

**Design**

Background: Solid black #000000 across the whole slide. Pure black #000000 background with no texture. Title in white #FFFFFF NVIDIASans Medium, sans-serif, about 4.9% of slide height, sentence case. Subtitle lines in grey #A1A1A1 NVIDIASans Regular about 2.0% of height, centred, with generous line spacing (about 4.1% pitch). Chart titles white NVIDIASans Medium about 2.0%; axis and data labels white NVIDIASans Regular about 1.5%; source line grey #757575. Accent colours: NVIDIA green #76B900 for Open-Weight and line series, blue #0074DF for Proprietary. Panels are outlined rounded rectangles with a thin grey stroke (about #5A5A5A, roughly 1 px) and no fill, corner radius about 1% of slide width, no shadow.

**Elements and positions** (back to front; x, y, width, height as % of the slide; type, colour and alignment in brackets)

- shape at x 0, y 0, w 100, h 100 [#000000]: Full-slide background rectangle (No stroke, no gradient, covers the slide.)
- shape at x 4.95, y 31, w 44.3, h 54.8 [#000000]: Left panel outline (rounded rectangle containing left chart) (No fill, 1px stroke about #5A5A5A, corner radius about 1% of slide width, no shadow.)
- shape at x 50.7, y 31, w 44.3, h 54.8 [#000000]: Right panel outline (rounded rectangle containing right chart) (No fill, 1px stroke about #5A5A5A, corner radius about 1% of slide width, no shadow.)
- title at x 11, y 10.5, w 78, h 4.9 [NVIDIASans Medium, sans-serif, sentence case; font size 4.94% of slide height; #FFFFFF; center aligned]: Open Models Enable New Builders and Create New Markets
- subtitle at x 27.2, y 18.1, w 45.9, h 2 [NVIDIASans Regular, sans-serif; font size 1.98% of slide height; #A1A1A1; center aligned]: Advancements in open models are contributing to the usefulness and proliferation of AI.
- body at x 18.9, y 22.2, w 62.6, h 2 [NVIDIASans Regular, sans-serif; font size 1.98% of slide height; #A1A1A1; center aligned]: Open-weight models run primarily on NVIDIA and account for ~75% of token generation today, up from ~40% a year ago.
- body at x 31.3, y 26.3, w 37.7, h 2 [NVIDIASans Regular, sans-serif; font size 1.98% of slide height; #A1A1A1; center aligned]: Token consumption is growing exponentially – up 25X over the past year.
- chart at x 10.2, y 41.3, w 36.7, h 35.3 [#76B900]: 100% stacked area chart, Open Router Token Volume - Open vs. Closed, with Proprietary (blue, bottom) and Open-Weight (green, top) series, monthly Jan-25 to Aug-26 (Plot area top at 100% (y 41.3) and bottom at 0% (y 76.6). Proprietary fill #0074DF from the baseline up to the about values listed in visuals; Open-Weight fill #76B900 from there to 100%. No gridlines, no outline, straight segments between monthly points, x labels rotated about 45 degrees.)
- shape at x 21.8, y 38, w 0.6, h 0.9 [#0074DF]: Legend marker, blue square for Proprietary (Small square swatch to the left of the Proprietary label.)
- shape at x 27.9, y 38, w 0.6, h 0.9 [#76B900]: Legend marker, green square for Open-Weight (Small square swatch to the left of the Open-Weight label.)
- caption at x 22.4, y 37.8, w 4.4, h 1.5 [NVIDIASans Regular, sans-serif; font size 1.48% of slide height; #FFFFFF; left aligned]: Proprietary
- caption at x 28.5, y 37.8, w 5.2, h 1.5 [NVIDIASans Regular, sans-serif; font size 1.48% of slide height; #FFFFFF; left aligned]: Open-Weight
- body at x 15.3, y 34.3, w 23.7, h 2 [NVIDIASans Medium, sans-serif; font size 1.98% of slide height; #FFFFFF; center aligned]: Open Router Token Volume - Open vs. Closed
- body at x 63.6, y 34.3, w 18.6, h 2 [NVIDIASans Medium, sans-serif; font size 1.98% of slide height; #FFFFFF; center aligned]: Open Router Monthly Token Traffic
- caption at x 71.1, y 37, w 3.5, h 1.5 [NVIDIASans Regular, sans-serif; font size 1.48% of slide height; #FFFFFF; center aligned]: (Trillions)
- chart at x 52.7, y 42.4, w 39, h 35.5 [#76B900]: Line chart, Open Router Monthly Token Traffic (Trillions), 20 monthly values from 2 to 379 with printed data labels (Line about 2.5pt #76B900 with no markers; white data labels above points; grey #757575 horizontal baseline at value 0 (y 76.6) running about x 56 to 91.8; no vertical axis line; no gridlines.)
- caption at x 52.7, y 42.4, w 1.7, h 34.2 [NVIDIASans Regular, sans-serif; font size 1.48% of slide height; #FFFFFF; right aligned]: Y axis labels 0, 50, 100, 150, 200, 250, 300, 350, 400 (right chart, trillions) (Stacked vertically at 50-unit spacing: 400 at y 42.4, 350 at y 46.6, 300 at y 50.7, 250 at y 54.9, 200 at y 59.1, 150 at y 63.3, 100 at y 67.5, 50 at y 71.7, 0 at y 75.9.)
- caption at x 7.2, y 40.6, w 2.2, h 35.3 [NVIDIASans Regular, sans-serif; font size 1.48% of slide height; #FFFFFF; right aligned]: Y axis labels 0%, 10%, 20%, 30%, 40%, 50%, 60%, 70%, 80%, 90%, 100% (left chart) (Stacked vertically at 10% spacing from 100% at y 40.6 down to 0% at y 75.9.)
- caption at x 10.2, y 77.7, w 36.7, h 6 [NVIDIASans Regular, sans-serif; font size 1.48% of slide height; #FFFFFF; right aligned]: X axis labels Jan-25 through Aug-26 (left chart), monthly, rotated about 45 degrees (Labels end at the tick positions beneath each month; rotated text reads upward to the right.)
- caption at x 56.2, y 66, w 36.5, h 8.4 [NVIDIASans Regular, sans-serif; font size 1.48% of slide height; #FFFFFF; center aligned]: Data labels 2 3 6 8 10 9 11 15 21 22 27 26 32 50 84 97 123 188 248 379 (right chart, above line points) (Each label sits just above its point on the green line; the final values climb steeply from 50 to 379.)
- caption at x 55.1, y 77.7, w 36.1, h 3.5 [NVIDIASans Regular, sans-serif; font size 1.48% of slide height; #FFFFFF; center aligned]: X axis labels Jan 2025, Apr 2025, Jul 2025, Oct 2025, Jan 2026, Apr 2026, Jul 2026 (right chart), two-line stacked month over year (Each label is two lines: month on top, year below; positioned at the quarterly ticks.)
- caption at x 86.2, y 82.9, w 7.9, h 1.5 [NVIDIASans Regular, sans-serif; font size 1.48% of slide height; #757575; right aligned]: Source: OpenRouter
- logo at x 92.9, y 94, w 4.6, h 2 [#76B900]: NVIDIA logo (green eye mark with white NVIDIA wordmark) (Bottom-right corner logo, green mark with white wordmark, no overlap with other elements.)

**Measured from the file** (exact, read from the PDF rather than estimated; x, y, w, h as % of the slide; font size as % of the slide height)

- fill at x 0.0, y 0.0, w 100.0, h 100.0: #000000 (covers the slide)
- fill at x 10.2, y 48.3, w 36.7, h 28.3: #0074DF
- fill at x 10.2, y 41.3, w 36.7, h 26.8: #76B900
- text at x 11.0, y 10.5, w 78.0, h 4.9 [NVIDIASans-Medium; font size 4.94; #FFFFFF]: Open Models Enable New Builders and Create New Markets
- text at x 27.2, y 18.1, w 45.9, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: Advancements in open models are contributing to the usefulness and proliferation of AI.
- text at x 18.9, y 22.2, w 62.6, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: Open-weight models run primarily on NVIDIA and account for ~75% of token generation today, up from ~40% a year ago.
- text at x 31.3, y 26.3, w 37.7, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: Token consumption is growing exponentially – up 25X over the past year.
- text at x 15.3, y 34.3, w 23.7, h 2.0 [NVIDIASans-Medium; font size 1.98; #FFFFFF]: Open Router Token Volume - Open vs. Closed
- text at x 63.6, y 34.3, w 18.6, h 2.0 [NVIDIASans-Medium; font size 1.98; #FFFFFF]: Open Router Monthly Token Traffic
- text at x 71.1, y 37.0, w 3.5, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: (Trillions)
- text at x 22.4, y 37.8, w 4.4, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Proprietary
- text at x 28.5, y 37.8, w 5.2, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Open-Weight
- text at x 7.2, y 40.6, w 2.2, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: 100%
- text at x 52.7, y 42.4, w 1.7, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: 400
- text at x 90.3, y 42.8, w 1.7, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: 379
- text at x 7.7, y 44.1, w 1.7, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: 90%
- text at x 52.7, y 46.6, w 1.7, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: 350
- text at x 7.7, y 47.6, w 1.7, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: 80%
- text at x 7.7, y 51.1, w 1.7, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: 70%
- text at x 52.7, y 50.7, w 1.7, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: 300
- text at x 87.3, y 53.9, w 1.7, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: 248
- text at x 7.7, y 54.7, w 1.7, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: 60%
- text at x 52.7, y 54.9, w 1.7, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: 250
- text at x 86.1, y 57.8, w 1.7, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: 188
- text at x 7.7, y 58.2, w 1.7, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: 50%
- text at x 52.7, y 59.1, w 1.7, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: 200
- text at x 7.7, y 61.7, w 1.7, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: 40%
- text at x 52.7, y 63.3, w 1.7, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: 150
- text at x 84.2, y 63.3, w 1.7, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: 123
- text at x 7.7, y 65.3, w 1.7, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: 30%
- text at x 83.1, y 66.0, w 1.2, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: 97
- text at x 52.7, y 67.5, w 1.7, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: 100
- text at x 81.2, y 67.1, w 1.2, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: 84
- text at x 7.7, y 68.8, w 1.7, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: 20%
- text at x 79.2, y 69.9, w 1.2, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: 50
- text at x 77.5, y 71.3, w 1.2, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: 32
- text at x 53.2, y 71.7, w 1.2, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: 50
- text at x 76.0, y 72.2, w 1.2, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: 26
- text at x 7.7, y 72.3, w 1.7, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: 10%
- text at x 70.6, y 72.2, w 4.9, h 2.0 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: 21 22 27
- text at x 63.4, y 73.7, w 1.2, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: 10
- text at x 67.0, y 73.1, w 3.0, h 1.9 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: 11 15
- text at x 56.2, y 74.4, w 0.7, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: 2
- text at x 58.2, y 74.3, w 0.7, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: 3
- text at x 60.0, y 74.1, w 0.7, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: 6
- text at x 61.8, y 73.9, w 0.7, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: 8
- text at x 65.2, y 73.7, w 0.7, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: 9
- text at x 8.2, y 75.9, w 1.2, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: 0%
- text at x 53.5, y 75.9, w 0.5, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: -
- text at x 53.7, y 76.6, w 0.5, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: 0
- text at x 55.4, y 77.7, w 1.4, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Jan
- text at x 61.1, y 77.7, w 1.4, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Apr
- text at x 66.8, y 77.7, w 1.2, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Jul
- text at x 72.3, y 77.7, w 1.4, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Oct
- text at x 77.9, y 77.7, w 1.4, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Jan
- text at x 83.5, y 77.7, w 1.4, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Apr
- text at x 89.3, y 77.7, w 1.2, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Jul
- text at x 55.1, y 79.3, w 2.0, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: 2025
- text at x 60.8, y 79.3, w 2.0, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: 2025
- text at x 66.4, y 79.3, w 2.0, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: 2025
- text at x 72.0, y 79.3, w 2.0, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: 2025
- text at x 77.6, y 79.3, w 2.0, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: 2026
- text at x 83.2, y 79.3, w 2.0, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: 2026
- text at x 88.9, y 79.3, w 2.0, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: 2026
- text at x 86.2, y 82.9, w 7.9, h 1.5 [NVIDIASans-Regular; font size 1.48; #757575]: Source: OpenRouter

### Slide 24

**Text**

NVIDIA is the Largest
Contributor to Open Models
NVIDIA builds frontier open models and contributes data, software, and standards
across the AI ecosystem. We are committed to investing in and growing the open
source community to enable the ecosystem and drive accelerated growth.
NVIDIA’s investments in open models span Nemotron (Language and Agentic AI),
Cosmos (Physical AI), Isaac GR00T (Robotics), Alpamayo (Autonomous Vehicles),
and more.
On September 3rd, NVIDIA announced its intent to acquire Hugging Face.
Together, we will scale Hugging Face’s platform, strengthen its infrastructure
and expand access to AI for developers and institutions worldwide.
Omniverse
Libraries & microservices
for developing
Physical AI applications
Cosmos 3
Open world
foundation model
Isaac GR00T
Open humanoid-robot
reference platform
BioNeMo
Proteina-Complexa
frontier generative
protein model
Alpamayo
The world’s first
reasoning autonomous
vehicle model
1,000
800
600
400
200
0
Repos
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep
2025-2026
NVIDIA
Alibaba Cloud
Hugging Face
Meta
Microsoft
IBM
Google
Baidu
AMD
ByteDance
Z.ai
Source: Hugging Face

**Layout**

Two columns on a black field: a left text column starting at about 27.5% of slide height with the title and three body paragraphs, left-aligned at x 5.2%; a right column holding a row of five outlined cards across the top (y 18.8% to 29.4%) above a single outlined chart panel (y 31.0% to 85.8%, x 50.7% to 95.0%). No footer text or slide number; the NVIDIA logo sits bottom right and the AI World logo sits inside the chart panel bottom right.

**Visuals**

Line chart inside a rounded outlined panel. Type: multi-series cumulative line chart titled by the y-axis label 'Repos'; x-axis labelled with months Jan to Sep spanning 2025-2026 (caption '2025-2026' below axis). Y-axis ticks printed: 0, 200, 400, 600, 800, 1,000 (range 0 to about 1,000). X-axis runs Jan 2025 to Sep 2026 (21 monthly ticks). Series (values printed only as end labels; all others are estimates): NVIDIA in bright green #76B900, thick line, highlighted with a green-outlined circle containing the NVIDIA logo and label 'NVIDIA', reaching about 950 by Sep 2026; Alibaba Cloud in orange #F26B1D, ending about 510 (label circled orange with a bracket icon); Hugging Face in yellow #FFD21E, ending about 380 (circle with a yellow face icon); Meta in blue #1877F2, ending about 290 (circle with infinity icon); Microsoft in red-orange, about 245 (small logo square); IBM in light blue, about 230; Google in green, about 215 (four-colour icon); Baidu in dark red-blue, about 190 (paw icon); AMD in red, about 180 (black circle with red arrow icon); ByteDance in cyan #25F4EE, about 135 (bars icon, flat line); Z.ai in grey, about 80 (circle with Z icon). Each end label has a small circular logo outline in its series colour. No legend box; labels act as legend. Highlighted point: NVIDIA end point circled. Source line 'Source: Hugging Face' with 'Hugging Face' underlined in green. AI World logo: blue square with snowflake-like icon and 'AI WORLD' text in blue, cropped as a placed image. Three card-row icons: none (cards are text only). NVIDIA logo bottom right in green and white.

**Design**

Background: Solid black #000000. Entirely dark: black background #000000, white titles #FFFFFF, body text grey #A1A1A1, card titles and accent green #76B900, source label grey #757575, chart panel border grey about #5A5A5A 1px. Typeface NVIDIASans, a geometric humanist sans-serif. Title Medium weight about 4.9% of slide height (cap height about 3.4%). Body Regular about 2.0% of slide height. Card titles Medium about 1.7%, card body Regular about 1.1%. Chart axis labels Medium about 1.0%. Cards: rounded rectangles with thin grey outline about #6B6B6B, corner radius about 1% of slide width, black fill, no shadow. Chart panel: rounded rectangle, corner radius about 1%, black fill. Lines are smooth stepped polylines with 1 to 3 px stroke. No gradients or shadows.

**Elements and positions** (back to front; x, y, width, height as % of the slide; type, colour and alignment in brackets)

- shape at x 0, y 0, w 100, h 100 [#000000]: Full-slide black background fill (Covers the slide; no outline.)
- shape at x 50.7, y 18.8, w 8.2, h 10.6 [#000000]: Card outline 1 (Omniverse) (Rounded rectangle, black fill, 1px grey outline about #6B6B6B, corner radius about 1%.)
- shape at x 59.8, y 18.8, w 8.2, h 10.6 [#000000]: Card outline 2 (Cosmos 3) (Rounded rectangle, black fill, 1px grey outline about #6B6B6B, corner radius about 1%.)
- shape at x 68.8, y 18.8, w 8.2, h 10.6 [#000000]: Card outline 3 (Isaac GR00T) (Rounded rectangle, black fill, 1px grey outline about #6B6B6B, corner radius about 1%.)
- shape at x 77.9, y 18.8, w 8.2, h 10.6 [#000000]: Card outline 4 (BioNeMo) (Rounded rectangle, black fill, 1px grey outline about #6B6B6B, corner radius about 1%.)
- shape at x 86.9, y 18.8, w 8.2, h 10.6 [#000000]: Card outline 5 (Alpamayo) (Rounded rectangle, black fill, 1px grey outline about #6B6B6B, corner radius about 1%.)
- text at x 52.3, y 21.2, w 4.9, h 1.7 [NVIDIASans Medium; font size 1.73% of slide height; #76B900; center aligned]: Omniverse
- text at x 51.1, y 23.5, w 7.4, h 4 [NVIDIASans Regular; font size 1.11% of slide height; #FFFFFF; center aligned]: Libraries & microservices / for developing / Physical AI applications (Three lines, line spacing about 1.4.)
- text at x 61.6, y 21.8, w 4.5, h 1.7 [NVIDIASans Medium; font size 1.73% of slide height; #76B900; center aligned]: Cosmos 3
- text at x 61.3, y 24.2, w 5.2, h 2.4 [NVIDIASans Regular; font size 1.11% of slide height; #FFFFFF; center aligned]: Open world / foundation model (Two lines, line spacing about 1.4.)
- text at x 70, y 21.8, w 5.7, h 1.7 [NVIDIASans Medium; font size 1.73% of slide height; #76B900; center aligned]: Isaac GR00T
- text at x 69.6, y 24.2, w 6.7, h 2.4 [NVIDIASans Regular; font size 1.11% of slide height; #FFFFFF; center aligned]: Open humanoid-robot / reference platform (Two lines, line spacing about 1.4.)
- text at x 79.8, y 21.2, w 4.2, h 1.7 [NVIDIASans Medium; font size 1.73% of slide height; #76B900; center aligned]: BioNeMo
- text at x 79.1, y 23.5, w 5.7, h 4 [NVIDIASans Regular; font size 1.11% of slide height; #FFFFFF; center aligned]: Proteina-Complexa / frontier generative / protein model (Three lines, line spacing about 1.4.)
- text at x 88.7, y 21.2, w 4.5, h 1.7 [NVIDIASans Medium; font size 1.73% of slide height; #76B900; center aligned]: Alpamayo
- text at x 88.6, y 23.5, w 6.9, h 4 [NVIDIASans Regular; font size 1.11% of slide height; #FFFFFF; center aligned]: The world’s first / reasoning autonomous / vehicle model (Three lines, line spacing about 1.4.)
- shape at x 50.7, y 31, w 44.4, h 54.8 [#000000]: Chart panel outline (Rounded rectangle, black fill, 1px grey outline about #5A5A5A, corner radius about 1%.)
- chart at x 52, y 35, w 41, h 45 [#76B900]: Cumulative line chart of open model repos, titled by y-axis label 'Repos', x-axis Jan 2025 to Sep 2026, y-axis 0 to 1,000; series NVIDIA (about 950 at Sep), Alibaba Cloud (about 510), Hugging Face (about 380), Meta (about 290), Microsoft (about 245), IBM (about 230), Google (about 215), Baidu (about 190), AMD (about 180), ByteDance (about 135), Z.ai (about 80) (NVIDIA line thick green #76B900; other lines 1 to 2px in orange #F26B1D, yellow #FFD21E, blue #1877F2, cyan #25F4EE, red, grey; plot area black; no gridlines; NVIDIA endpoint circled by green ring.)
- text at x 53, y 35.8, w 1.5, h 1 [NVIDIASans Medium; font size 0.99% of slide height; #FFFFFF; right aligned]: 1,000
- text at x 53.5, y 43.9, w 1, h 1 [NVIDIASans Medium; font size 0.99% of slide height; #FFFFFF; right aligned]: 800
- text at x 51.2, y 57.5, w 2, h 1.2 [NVIDIASans Medium; font size 1.24% of slide height; #FFFFFF; left aligned]: Repos (Y-axis title.)
- text at x 53.5, y 51.9, w 1, h 1 [NVIDIASans Medium; font size 0.99% of slide height; #FFFFFF; right aligned]: 600
- text at x 53.5, y 60, w 1, h 1 [NVIDIASans Medium; font size 0.99% of slide height; #FFFFFF; right aligned]: 400
- text at x 53.5, y 68, w 1, h 1 [NVIDIASans Medium; font size 0.99% of slide height; #FFFFFF; right aligned]: 200
- text at x 54.2, y 76.1, w 0.3, h 1 [NVIDIASans Medium; font size 0.99% of slide height; #FFFFFF; right aligned]: 0
- text at x 54.3, y 77.3, w 1, h 1 [NVIDIASans Medium; font size 0.99% of slide height; #FFFFFF; center aligned]: Jan (X-axis tick label.)
- text at x 56, y 77.3, w 7.6, h 1 [NVIDIASans Medium; font size 0.99% of slide height; #FFFFFF; center aligned]: Feb Mar Apr May Jun (X-axis tick labels.)
- text at x 64.3, y 77.3, w 0.8, h 1 [NVIDIASans Medium; font size 0.99% of slide height; #FFFFFF; center aligned]: Jul (X-axis tick label.)
- text at x 65.9, y 77.3, w 2.7, h 1 [NVIDIASans Medium; font size 0.99% of slide height; #FFFFFF; center aligned]: Aug Sep (X-axis tick labels.)
- text at x 69.2, y 77.3, w 1, h 1 [NVIDIASans Medium; font size 0.99% of slide height; #FFFFFF; center aligned]: Oct (X-axis tick label.)
- text at x 70.8, y 77.3, w 4.3, h 1 [NVIDIASans Medium; font size 0.99% of slide height; #FFFFFF; center aligned]: Nov Dec Jan (X-axis tick labels.)
- text at x 75.8, y 77.3, w 7.6, h 1 [NVIDIASans Medium; font size 0.99% of slide height; #FFFFFF; center aligned]: Feb Mar Apr May Jun (X-axis tick labels.)
- text at x 84.2, y 77.3, w 0.8, h 1 [NVIDIASans Medium; font size 0.99% of slide height; #FFFFFF; center aligned]: Jul (X-axis tick label.)
- text at x 85.7, y 77.3, w 2.7, h 1 [NVIDIASans Medium; font size 0.99% of slide height; #FFFFFF; center aligned]: Aug Sep (X-axis tick labels.)
- text at x 70.3, y 79.9, w 3.6, h 1.2 [NVIDIASans Medium; font size 1.24% of slide height; #FFFFFF; center aligned]: 2025-2026 (X-axis title.)
- logo at x 87.6, y 36, w 2.6, h 4 [#76B900]: NVIDIA logo in green circle outline, endpoint label for NVIDIA series (Circle outline 2px #76B900 with NVIDIA eye logo in green; label 'NVIDIA' text beside it.)
- text at x 90.7, y 37.8, w 3.2, h 1.6 [NVIDIASans Bold; font size 1.61% of slide height; #FFFFFF; left aligned]: NVIDIA
- logo at x 87.6, y 54.8, w 1.6, h 2.4 [#F26B1D]: Alibaba Cloud logo circle (Circle outline orange #F26B1D with bracket icon.)
- text at x 89.8, y 55.2, w 2.4, h 2.4 [NVIDIASans Medium; font size 1.11% of slide height; #FFFFFF; left aligned]: Alibaba / Cloud (Two lines.)
- logo at x 87.6, y 59.4, w 1.6, h 2.4 [#FFD21E]: Hugging Face logo circle (Circle outline yellow #FFD21E with yellow face icon.)
- text at x 89.8, y 60, w 2.7, h 2.4 [NVIDIASans Medium; font size 1.11% of slide height; #FFFFFF; left aligned]: Hugging / Face (Two lines.)
- logo at x 87.6, y 63.9, w 1.6, h 2.4 [#1877F2]: Meta logo circle (Circle outline blue #1877F2 with infinity icon.)
- text at x 89.8, y 64.4, w 1.5, h 1.1 [NVIDIASans Medium; font size 1.11% of slide height; #FFFFFF; left aligned]: Meta
- logo at x 87.6, y 65.3, w 1.6, h 1.8 [#F25022]: Microsoft logo circle (Small circle outline with four-square Microsoft icon.)
- text at x 89.8, y 65.6, w 2.3, h 0.9 [NVIDIASans Medium; font size 0.87% of slide height; #FFFFFF; left aligned]: Microsoft
- logo at x 87.6, y 66.1, w 1.6, h 1.6 [#1F70C1]: IBM logo circle (Small circle outline with IBM mark.)
- text at x 89.8, y 66.4, w 0.9, h 0.9 [NVIDIASans Medium; font size 0.87% of slide height; #FFFFFF; left aligned]: IBM
- logo at x 87.6, y 66.9, w 1.6, h 1.6 [#34A853]: Google logo circle (Small circle outline with Google G icon.)
- text at x 89.8, y 67.2, w 1.6, h 0.9 [NVIDIASans Medium; font size 0.87% of slide height; #FFFFFF; left aligned]: Google
- logo at x 87.6, y 68, w 1.6, h 1.6 [#2932E1]: Baidu logo circle (Small circle outline with Baidu icon.)
- text at x 89.8, y 68.2, w 1.3, h 0.9 [NVIDIASans Medium; font size 0.87% of slide height; #FFFFFF; left aligned]: Baidu
- logo at x 87.6, y 68.8, w 1.6, h 1.6 [#ED1C24]: AMD logo circle (Small circle outline with AMD arrow icon.)
- text at x 89.8, y 69, w 1.1, h 0.9 [NVIDIASans Medium; font size 0.87% of slide height; #FFFFFF; left aligned]: AMD
- logo at x 87.6, y 70, w 1.6, h 1.8 [#25F4EE]: ByteDance logo circle (Circle outline cyan #25F4EE with bars icon.)
- text at x 89.8, y 70.6, w 2.5, h 0.9 [NVIDIASans Medium; font size 0.87% of slide height; #FFFFFF; left aligned]: ByteDance
- logo at x 87.6, y 72, w 1.6, h 1.8 [#FFFFFF]: Z.ai logo circle (Circle outline grey-white with Z icon.)
- text at x 89.7, y 72.6, w 0.8, h 0.9 [NVIDIASans Medium; font size 0.87% of slide height; #FFFFFF; left aligned]: Z.ai
- text at x 52, y 82.2, w 2.3, h 1.1 [NVIDIASans Regular; font size 1.11% of slide height; #757575; left aligned]: Source:
- text at x 54.4, y 82.2, w 4.1, h 1.1 [NVIDIASans Medium; font size 1.11% of slide height; #76B900; left aligned]: Hugging Face (Underlined hyperlink.)
- image at x 89.5, y 81.3, w 4.3, h 2.5 [#1F4FD8]: AI World logo (blue square icon with 'AI WORLD' wordmark) (Placed image 258 by 83 pixels; blue square icon with white snowflake-like mark and blue 'AI WORLD' text.)
- text at x 5.2, y 27.5, w 37, h 10.2 [NVIDIASans Medium; font size 4.94% of slide height; #FFFFFF; left aligned]: NVIDIA is the Largest / Contributor to Open Models (Title, two lines, line spacing about 1.15.)
- text at x 5.2, y 41, w 42.8, h 8.9 [NVIDIASans Regular; font size 1.98% of slide height; #A1A1A1; left aligned]: NVIDIA builds frontier open models and contributes data, software, and standards / across the AI ecosystem. We are committed to investing in and growing the open / source community to enable the ecosystem and drive accelerated growth. (Paragraph, line spacing about 1.6.)
- text at x 5.2, y 51.4, w 42, h 8 [NVIDIASans Regular; font size 1.98% of slide height; #A1A1A1; left aligned]: NVIDIA’s investments in open models span Nemotron (Language and Agentic AI), / Cosmos (Physical AI), Isaac GR00T (Robotics), Alpamayo (Autonomous Vehicles), / and more. (Paragraph, line spacing about 1.6.)
- text at x 5.2, y 61.7, w 40.2, h 8 [NVIDIASans Regular; font size 1.98% of slide height; #A1A1A1; left aligned]: On September 3rd, NVIDIA announced its intent to acquire Hugging Face. / Together, we will scale Hugging Face’s platform, strengthen its infrastructure / and expand access to AI for developers and institutions worldwide. (Paragraph, line spacing about 1.6; 'rd' is superscript at about 1.3% em.)
- logo at x 92.9, y 94.2, w 4.6, h 1.6 [#76B900; right aligned]: NVIDIA logo (green eye mark with NVIDIA wordmark) (Bottom-right corner logo, white wordmark with green eye icon.)

**Measured from the file** (exact, read from the PDF rather than estimated; x, y, w, h as % of the slide; font size as % of the slide height)

- fill at x 0.0, y 0.0, w 100.0, h 100.0: #000000 (covers the slide)
- image at x 89.5, y 81.3, w 4.3, h 2.5: 258 by 83 pixels
- text at x 5.2, y 27.5, w 28.9, h 4.9 [NVIDIASans-Medium; font size 4.94; #FFFFFF]: NVIDIA is the Largest
- text at x 5.2, y 32.8, w 37.0, h 4.9 [NVIDIASans-Medium; font size 4.94; #FFFFFF]: Contributor to Open Models
- text at x 52.3, y 21.2, w 4.9, h 1.7 [NVIDIASans-Medium; font size 1.73; #76B900]: Omniverse
- text at x 61.6, y 21.8, w 4.5, h 1.7 [NVIDIASans-Medium; font size 1.73; #76B900]: Cosmos 3
- text at x 70.0, y 21.8, w 5.7, h 1.7 [NVIDIASans-Medium; font size 1.73; #76B900]: Isaac GR00T
- text at x 79.8, y 21.2, w 4.2, h 1.7 [NVIDIASans-Medium; font size 1.73; #76B900]: BioNeMo
- text at x 88.7, y 21.2, w 4.5, h 1.7 [NVIDIASans-Medium; font size 1.73; #76B900]: Alpamayo
- text at x 5.2, y 41.0, w 42.8, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: NVIDIA builds frontier open models and contributes data, software, and standards
- text at x 51.1, y 23.5, w 7.4, h 1.1 [NVIDIASans-Regular; font size 1.11; #FFFFFF]: Libraries & microservices
- text at x 79.1, y 23.5, w 5.7, h 1.1 [NVIDIASans-Regular; font size 1.11; #FFFFFF]: Proteina-Complexa
- text at x 88.6, y 23.5, w 4.9, h 1.1 [NVIDIASans-Regular; font size 1.11; #FFFFFF]: The world’s first
- text at x 5.2, y 43.9, w 42.3, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: across the AI ecosystem. We are committed to investing in and growing the open
- text at x 62.2, y 24.2, w 3.5, h 1.1 [NVIDIASans-Regular; font size 1.11; #FFFFFF]: Open world
- text at x 69.6, y 24.2, w 6.7, h 1.1 [NVIDIASans-Regular; font size 1.11; #FFFFFF]: Open humanoid-robot
- text at x 52.7, y 24.9, w 4.3, h 1.1 [NVIDIASans-Regular; font size 1.11; #FFFFFF]: for developing
- text at x 79.2, y 24.9, w 5.6, h 1.1 [NVIDIASans-Regular; font size 1.11; #FFFFFF]: frontier generative
- text at x 87.6, y 24.9, w 6.9, h 1.1 [NVIDIASans-Regular; font size 1.11; #FFFFFF]: reasoning autonomous
- text at x 61.3, y 25.5, w 5.2, h 1.1 [NVIDIASans-Regular; font size 1.11; #FFFFFF]: foundation model
- text at x 70.1, y 25.5, w 5.5, h 1.1 [NVIDIASans-Regular; font size 1.11; #FFFFFF]: reference platform
- text at x 5.2, y 46.9, w 38.6, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: source community to enable the ecosystem and drive accelerated growth.
- text at x 51.4, y 26.2, w 6.8, h 1.1 [NVIDIASans-Regular; font size 1.11; #FFFFFF]: Physical AI applications
- text at x 79.9, y 26.2, w 4.0, h 1.1 [NVIDIASans-Regular; font size 1.11; #FFFFFF]: protein model
- text at x 89.0, y 26.2, w 4.0, h 1.1 [NVIDIASans-Regular; font size 1.11; #FFFFFF]: vehicle model
- text at x 90.7, y 37.8, w 3.2, h 1.6 [NVIDIASans-Bold; font size 1.61; #FFFFFF]: NVIDIA
- text at x 5.2, y 51.4, w 42.0, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: NVIDIA’s investments in open models span Nemotron (Language and Agentic AI),
- text at x 5.2, y 54.3, w 41.4, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: Cosmos (Physical AI), Isaac GR00T (Robotics), Alpamayo (Autonomous Vehicles),
- text at x 5.2, y 57.3, w 5.4, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: and more.
- text at x 5.2, y 61.7, w 37.9, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: On September 3 , NVIDIA announced its intent to acquire Hugging Face.
- text at x 5.3, y 64.7, w 40.2, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: Together, we will scale Hugging Face’s platform, strengthen its infrastructure
- text at x 5.3, y 67.6, w 34.8, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: and expand access to AI for developers and institutions worldwide.
- text at x 53.0, y 35.8, w 1.5, h 1.0 [NVIDIASans-Medium; font size 0.99; #FFFFFF]: 1,000
- text at x 53.5, y 43.9, w 1.0, h 1.0 [NVIDIASans-Medium; font size 0.99; #FFFFFF]: 800
- text at x 51.2, y 57.5, w 2.0, h 1.2 [NVIDIASans-Medium; font size 1.24; #FFFFFF]: Repos
- text at x 13.7, y 61.7, w 0.7, h 1.3 [NVIDIASans-Regular; font size 1.32; #A1A1A1]: rd
- text at x 89.8, y 55.2, w 2.4, h 1.1 [NVIDIASans-Medium; font size 1.11; #FFFFFF]: Alibaba
- text at x 89.8, y 56.5, w 1.7, h 1.1 [NVIDIASans-Medium; font size 1.11; #FFFFFF]: Cloud
- text at x 53.5, y 51.9, w 1.0, h 1.0 [NVIDIASans-Medium; font size 0.99; #FFFFFF]: 600
- text at x 89.8, y 60.0, w 2.7, h 1.1 [NVIDIASans-Medium; font size 1.11; #FFFFFF]: Hugging
- text at x 89.8, y 61.3, w 1.4, h 1.1 [NVIDIASans-Medium; font size 1.11; #FFFFFF]: Face
- text at x 89.8, y 64.4, w 1.5, h 1.1 [NVIDIASans-Medium; font size 1.11; #FFFFFF]: Meta
- text at x 53.5, y 60.0, w 1.0, h 1.0 [NVIDIASans-Medium; font size 0.99; #FFFFFF]: 400
- text at x 70.3, y 79.9, w 3.6, h 1.2 [NVIDIASans-Medium; font size 1.24; #FFFFFF]: 2025-2026
- text at x 53.5, y 68.0, w 1.0, h 1.0 [NVIDIASans-Medium; font size 0.99; #FFFFFF]: 200
- text at x 52.0, y 82.2, w 2.3, h 1.1 [NVIDIASans-Regular; font size 1.11; #757575]: Source:
- text at x 54.4, y 82.2, w 4.1, h 1.1 [NVIDIASans-Medium; font size 1.11; #76B900]: Hugging Face
- text at x 89.8, y 65.6, w 2.3, h 0.9 [NVIDIASans-Medium; font size 0.87; #FFFFFF]: Microsoft
- text at x 89.8, y 66.4, w 0.9, h 0.9 [NVIDIASans-Medium; font size 0.87; #FFFFFF]: IBM
- text at x 54.2, y 76.1, w 0.3, h 1.0 [NVIDIASans-Medium; font size 0.99; #FFFFFF]: 0
- text at x 89.8, y 67.2, w 1.6, h 0.9 [NVIDIASans-Medium; font size 0.87; #FFFFFF]: Google
- text at x 54.3, y 77.3, w 1.0, h 1.0 [NVIDIASans-Medium; font size 0.99; #FFFFFF]: Jan
- text at x 56.0, y 77.3, w 7.6, h 1.0 [NVIDIASans-Medium; font size 0.99; #FFFFFF]: Feb Mar Apr May Jun
- text at x 64.3, y 77.3, w 0.8, h 1.0 [NVIDIASans-Medium; font size 0.99; #FFFFFF]: Jul
- text at x 65.9, y 77.3, w 2.7, h 1.0 [NVIDIASans-Medium; font size 0.99; #FFFFFF]: Aug Sep
- text at x 69.2, y 77.3, w 1.0, h 1.0 [NVIDIASans-Medium; font size 0.99; #FFFFFF]: Oct
- text at x 70.8, y 77.3, w 4.3, h 1.0 [NVIDIASans-Medium; font size 0.99; #FFFFFF]: Nov Dec Jan
- text at x 75.8, y 77.3, w 7.6, h 1.0 [NVIDIASans-Medium; font size 0.99; #FFFFFF]: Feb Mar Apr May Jun
- text at x 84.2, y 77.3, w 0.8, h 1.0 [NVIDIASans-Medium; font size 0.99; #FFFFFF]: Jul
- text at x 85.7, y 77.3, w 2.7, h 1.0 [NVIDIASans-Medium; font size 0.99; #FFFFFF]: Aug Sep
- text at x 89.8, y 68.2, w 1.3, h 0.9 [NVIDIASans-Medium; font size 0.87; #FFFFFF]: Baidu
- text at x 89.8, y 69.0, w 1.1, h 0.9 [NVIDIASans-Medium; font size 0.87; #FFFFFF]: AMD
- text at x 89.8, y 70.6, w 2.5, h 0.9 [NVIDIASans-Medium; font size 0.87; #FFFFFF]: ByteDance
- text at x 89.7, y 72.6, w 0.8, h 0.9 [NVIDIASans-Medium; font size 0.87; #FFFFFF]: Z.ai

### Slide 25

**Text**

NVIDIA Nemotron at the Frontier
Nemotron models are the most open, post-trainable, and efficient U.S. open-weight
models on the Artificial Analysis Intelligence Index.
Delivers intelligence at ~5X faster output speed and ~30% lower cost to task
completion than the best alternative open model.
NVIDIA releases weights, training data, and recipes, so enterprises, governments,
and startups can specialize models on their own data and own the result.
Nemotron adoption is broad and accelerating across Enterprise SaaS, AI Natives/
Startups, and Sovereign.
4 out of 5 top Enterprise SaaS Platforms have deployed Nemotron models.
AI Natives and post-training partners are fine-tuning Nemotron across industries
and domains.
MODEL EVALUATION
Breaking the Performance/Cost Frontier
Performance to API Price
90%
80%
70%
60%
50%
40%
30%
20%
Traj. Nemotron
3 Ultra
Traj. Nemotron
3 Super
Sonnet 4.6
GPT 5.5
Nemotron
3 Ultra
Nemotron
3 Super
GPT-OSS 120B
$1
$4
$16
Rubric Pass Rate
Price per 1M tokens

**Layout**

Two-column split on a black slide. The left column is a left-aligned text block, vertically centred, with the title at the top of the block and six body paragraphs below it, starting at about x 5.2% and running to about x 48.5%. The right column holds a light-cream rounded panel (x 50.7% to 95.0%, y 18.9% to 85.8%) containing the chart image, inset about 0.6% horizontally and about 8.5% from the panel top. The NVIDIA logo sits alone in the bottom-right corner. There is no slide number or footer text.

**Visuals**

One chart, a scatter plot titled 'Breaking the Performance/Cost Frontier', with eyebrow 'MODEL EVALUATION' above the title and subtitle 'Performance to API Price' below it. Y axis: 'Rubric Pass Rate', linear, ticks at 20%, 30%, 40%, 50%, 60%, 70%, 80%, 90% (no unit label beyond percent). X axis: 'Price per 1M tokens', logarithmic-looking spacing, labelled ticks $1, $4, $16, with the plot starting at about $0.1 on the left. Points, with values estimated (about) because the slide prints none: Traj. Nemotron 3 Super is a light blue-violet bubble (#B4B9F5 with dark outline) at about 67% and about $0.2, with a fading vertical trail of smaller translucent blue-violet bubbles down to about 50%. Traj. Nemotron 3 Ultra is a peach bubble (#F5C9A6 with dark outline) at about 82% and about $1.0, with a fading trail of translucent peach bubbles down to about 62%. Nemotron 3 Super is a solid grey point (#9A9A9A) at about 43% and about $0.2, labelled 'Nemotron / 3 Super'. Nemotron 3 Ultra is a solid grey point at about 52% and about $1.0, labelled 'Nemotron / 3 Ultra'. GPT-OSS 120B is a grey point at about 22% and about $0.1, labelled 'GPT-OSS 120B'. GPT 5.5 is a grey point at about 78% and about $5, labelled 'GPT 5.5'. Sonnet 4.6 is a grey point at about 85% and about $10, labelled 'Sonnet 4.6'. A dashed dark-grey frontier curve (#6B6B6B) rises from about 25% at the left edge to about 86% at the right edge, passing near the Nemotron 3 Ultra point and the GPT 5.5 point. Gridlines are light grey on a cream background (#F6F9E8); axes are dark grey lines. Labels are in a black serif face, bold for the two Traj. Nemotron titles. No legend. Highlighted points are the two Traj. Nemotron bubbles, in blue-violet and peach.

**Design**

Background: Solid black #000000 covering the full slide. Full black background (#000000). Title in NVIDIA Sans Medium, white #FFFFFF, sentence case, about 4.9% of slide height. Body in NVIDIA Sans Regular, grey #A1A1A1, about 2.0% of slide height, line spacing about 1.5 lines, paragraphs separated by about 1.5% of slide height. Accent is the NVIDIA green (#76B900) in the logo only. The chart panel is a cream rectangle #F6F9E8 with a thin light-grey border, slightly rounded corners (about 0.3% radius) and no visible shadow. Chart text uses a bold sans for the title and a serif for axis and point labels.

**Elements and positions** (back to front; x, y, width, height as % of the slide; type, colour and alignment in brackets)

- shape at x 0, y 0, w 100, h 100 [#000000]: Full-slide background rectangle
- shape at x 50.7, y 18.9, w 44.3, h 66.9 [#F6F9E8]: Chart panel, rounded rectangle with thin light-grey border (corner radius about 0.3% of slide width; thin border #D9D9D9; no shadow; sits behind the chart image)
- chart at x 51.3, y 27.4, w 42.6, h 51.8 [#F6F9E8]: Scatter chart 'Breaking the Performance/Cost Frontier' with eyebrow MODEL EVALUATION and subtitle Performance to API Price; axes Rubric Pass Rate (20% to 90%) and Price per 1M tokens ($1, $4, $16); Traj. Nemotron 3 Super and 3 Ultra highlighted bubbles with trails; grey points for Nemotron 3 Super, Nemotron 3 Ultra, GPT-OSS 120B, GPT 5.5, Sonnet 4.6; dashed frontier curve (placed raster image 1302 by 895 pixels, boxed, not cropped, no overlay; its own cream background matches the panel)
- title at x 5.2, y 32.8, w 43.3, h 4.9 [NVIDIASans Medium, sentence case, normal tracking; font size 4.94% of slide height; #FFFFFF; left aligned]: NVIDIA Nemotron at the Frontier
- body at x 5.2, y 41, w 43.4, h 5 [NVIDIASans Regular, normal weight, sentence case; font size 1.98% of slide height; #A1A1A1; left aligned]: Nemotron models are the most open, post-trainable, and efficient U.S. open-weight / models on the Artificial Analysis Intelligence Index. (two lines, line pitch about 2.9% of slide height)
- body at x 5.3, y 48.4, w 40, h 5 [NVIDIASans Regular, normal weight, sentence case; font size 1.98% of slide height; #A1A1A1; left aligned]: Delivers intelligence at ~5X faster output speed and ~30% lower cost to task / completion than the best alternative open model. (two lines, line pitch about 2.9% of slide height)
- body at x 5.3, y 55.8, w 42.1, h 5 [NVIDIASans Regular, normal weight, sentence case; font size 1.98% of slide height; #A1A1A1; left aligned]: NVIDIA releases weights, training data, and recipes, so enterprises, governments, / and startups can specialize models on their own data and own the result. (two lines, line pitch about 2.9% of slide height)
- body at x 5.3, y 63.2, w 42, h 5 [NVIDIASans Regular, normal weight, sentence case; font size 1.98% of slide height; #A1A1A1; left aligned]: Nemotron adoption is broad and accelerating across Enterprise SaaS, AI Natives/ / Startups, and Sovereign. (two lines, line pitch about 2.9% of slide height)
- body at x 5.3, y 70.6, w 38.6, h 2 [NVIDIASans Regular, normal weight, sentence case; font size 1.98% of slide height; #A1A1A1; left aligned]: 4 out of 5 top Enterprise SaaS Platforms have deployed Nemotron models.
- body at x 5.3, y 75.1, w 42.3, h 4.9 [NVIDIASans Regular, normal weight, sentence case; font size 1.98% of slide height; #A1A1A1; left aligned]: AI Natives and post-training partners are fine-tuning Nemotron across industries / and domains. (two lines, line pitch about 2.9% of slide height)
- logo at x 92.9, y 94.2, w 4.6, h 1.6 [#76B900]: NVIDIA logo: green eye-shaped mark (#76B900) followed by white NVIDIA wordmark in bold sans, bottom-right corner (estimated box; not measured)

**Measured from the file** (exact, read from the PDF rather than estimated; x, y, w, h as % of the slide; font size as % of the slide height)

- fill at x 0.0, y 0.0, w 100.0, h 100.0: #000000 (covers the slide)
- fill at x 50.7, y 18.9, w 44.3, h 66.9: #F6F9E8
- image at x 51.3, y 27.4, w 42.6, h 51.8: 1302 by 895 pixels
- text at x 5.2, y 32.8, w 43.3, h 4.9 [NVIDIASans-Medium; font size 4.94; #FFFFFF]: NVIDIA Nemotron at the Frontier
- text at x 5.2, y 41.0, w 43.4, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: Nemotron models are the most open, post-trainable, and efficient U.S. open-weight
- text at x 5.3, y 43.9, w 26.2, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: models on the Artificial Analysis Intelligence Index.
- text at x 5.3, y 48.4, w 40.0, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: Delivers intelligence at ~5X faster output speed and ~30% lower cost to task
- text at x 5.3, y 51.4, w 25.6, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: completion than the best alternative open model.
- text at x 5.3, y 55.8, w 42.1, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: NVIDIA releases weights, training data, and recipes, so enterprises, governments,
- text at x 5.3, y 58.8, w 37.7, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: and startups can specialize models on their own data and own the result.
- text at x 5.3, y 63.2, w 42.0, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: Nemotron adoption is broad and accelerating across Enterprise SaaS, AI Natives/
- text at x 5.3, y 66.2, w 12.7, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: Startups, and Sovereign.
- text at x 5.3, y 70.6, w 38.6, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: 4 out of 5 top Enterprise SaaS Platforms have deployed Nemotron models.
- text at x 5.3, y 75.1, w 42.3, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: AI Natives and post-training partners are fine-tuning Nemotron across industries
- text at x 5.3, y 78.0, w 6.9, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: and domains.

### Slide 26

**Text**

Future Arriving
Space
Robotaxis
Robotics
Telco
Apptronik

**Layout**

Four equal portrait image columns with green category labels centred above each, on a black field; the title is centred near the top, and the NVIDIA logo sits at bottom right. Columns are evenly spaced with margins of about 5% left and right.

**Visuals**

Four portrait photographs in rounded boxes, left to right. (1) Space: Earth from orbit with a small satellite at upper left, a bright gold-yellow atmospheric limb arc along the upper right, city lights glowing across the dark night side at lower left, black space above; cool blue and white tones. (2) Robotaxis: a light silver-white electric sedan viewed from front three-quarter, with a woman passenger in pink top visible through the windshield, on a city street with trees, apartment buildings, two cyclists and parked cars behind; blue radar-like arcs drawn on the road, green lane tint at left and right, bright white road markings. (3) Robotics: a white and grey humanoid robot with a rounded head and two camera eyes, standing behind a white cylindrical home robot on a wooden table next to a clear drinking glass; kitchen background with grey cabinetry and a range hood; small 'Apptronik' caption at lower right. (4) Telco: a tall steel lattice telecom tower with multiple antenna panels, against blue sky with white clouds, rolling green forested hills and a small red-roofed town below; bright daylight. All photographs are boxed with thin rounded edges, not cropped to bleed.

**Design**

Background: Solid black #000000.

**Elements and positions** (back to front; x, y, width, height as % of the slide; type, colour and alignment in brackets)

- shape at x 0, y 0, w 100, h 100 [#000000]: Full-slide background fill (Covers the entire slide.)
- title at x 40.2, y 10.5, w 19.6, h 4.9 [NVIDIA Sans, Medium weight, sans-serif, mixed case; font size 4.94% of slide height; #FFFFFF; center aligned]: Future Arriving
- body at x 14, y 27, w 3.6, h 2.2 [NVIDIA Sans, Medium weight, sans-serif, mixed case; font size 2.22% of slide height; #76B900; center aligned]: Space (Column label centred over photo 1.)
- photo at x 5, y 31, w 21.7, h 54.8: Space photograph: Earth from orbit with satellite and city lights on the night side (Portrait boxed photo, thin grey-white rounded border, corner radius about 1%, not full-bleed; dark space at top, Earth with gold atmospheric glow lower half.)
- body at x 35.7, y 27, w 5.9, h 2.2 [NVIDIA Sans, Medium weight, sans-serif, mixed case; font size 2.22% of slide height; #76B900; center aligned]: Robotaxis (Column label centred over photo 2.)
- photo at x 27.8, y 31, w 21.6, h 54.8: Robotaxi photograph: silver electric sedan with passenger on city street, blue sensor arcs on road (Portrait boxed photo with rounded corners about 1%; not full-bleed; green lane tint and blue arc overlays on road.)
- body at x 58.8, y 27, w 5.2, h 2.2 [NVIDIA Sans, Medium weight, sans-serif, mixed case; font size 2.22% of slide height; #76B900; center aligned]: Robotics (Column label centred over photo 3.)
- photo at x 50.6, y 31, w 21.6, h 54.8: Robotics photograph: white humanoid robot beside white home robot on wooden table with glass (Portrait boxed photo with rounded corners about 1%; not full-bleed.)
- caption at x 68.4, y 83.7, w 3.2, h 1.2 [NVIDIA Sans, Regular weight, sans-serif, mixed case; font size 1.24% of slide height; #FFFFFF; right aligned]: Apptronik (Photo credit overlaid on lower right of robotics photo.)
- body at x 82.5, y 27, w 3.1, h 2.2 [NVIDIA Sans, Medium weight, sans-serif, mixed case; font size 2.22% of slide height; #76B900; center aligned]: Telco (Column label centred over photo 4.)
- photo at x 73.4, y 31, w 21.6, h 54.8: Telco photograph: steel lattice telecom tower with antennas over forested hills and town (Portrait boxed photo with rounded corners about 1%; not full-bleed.)
- logo at x 92.5, y 94.5, w 4.5, h 1.8 [#76B900]: NVIDIA logo (green eye mark with NVIDIA wordmark) (Bottom-right corner logo, green mark and white wordmark.)

**Measured from the file** (exact, read from the PDF rather than estimated; x, y, w, h as % of the slide; font size as % of the slide height)

- fill at x 0.0, y 0.0, w 100.0, h 100.0: #000000 (covers the slide)
- image at x 5.0, y 31.0, w 21.7, h 54.8: 450 by 640 pixels
- image at x 73.4, y 31.0, w 21.6, h 54.8: 1024 by 1452 pixels
- image at x 27.8, y 31.0, w 21.6, h 54.8: 1855 by 2644 pixels
- image at x 50.6, y 31.0, w 21.6, h 54.8: 466 by 665 pixels
- text at x 40.2, y 10.5, w 19.6, h 4.9 [NVIDIASans-Medium; font size 4.94; #FFFFFF]: Future Arriving
- text at x 14.0, y 27.0, w 3.6, h 2.2 [NVIDIASans-Medium; font size 2.22; #76B900]: Space
- text at x 35.7, y 27.0, w 5.9, h 2.2 [NVIDIASans-Medium; font size 2.22; #76B900]: Robotaxis
- text at x 58.8, y 27.0, w 5.2, h 2.2 [NVIDIASans-Medium; font size 2.22; #76B900]: Robotics
- text at x 82.5, y 27.0, w 3.1, h 2.2 [NVIDIASans-Medium; font size 2.22; #76B900]: Telco
- text at x 68.4, y 83.7, w 3.2, h 1.2 [NVIDIASans-Regular; font size 1.24; #FFFFFF]: Apptronik

### Slide 27

**Text**

Physical AI Adds a New Workload on AI Factories
Physical AI extends AI into the physical economy – manufacturing, robotics, autonomous vehicles and heavy industry.
It addresses a $50 trillion market and represents a multi-trillion dollar long-term opportunity for NVIDIA.
Only NVIDIA delivers all three computers necessary to build physical AI – Train (DGX), Simulate (Omniverse + Cosmos), and Run (Jetson Thor).
NVIDIA Physical AI Revenue ($ Billions)
Auto
Healthcare
Manufacturing
Smart Spaces
Telco
Robotics
20.0
16.0
12.0
8.0
4.0
0.0
2.9
3.3
5.2
9.3
8.2
10.8
(1H)
1H: $10.8B YTD
Already
$2.6B above
full-year FY26
FY22
FY23
FY24
FY25
FY26
FY27
Training
DGX
Simulation
Omniverse / Cosmos
Model Deployment
Jetson / IGX

**Layout**

Full-bleed black slide with a centred title and three centred subtitle lines across the top. Below, two equal rounded-corner panels sit side by side: a left diagram panel with three circular illustrations in a triangle, and a right bar chart panel. The NVIDIA logo sits bottom right; no slide number is visible.

**Visuals**

Left panel: a triangular cycle diagram. Top-left circle (Training, DGX) holds a white line sketch of server racks with gold-tinted highlights. Top-right circle (Simulation, Omniverse / Cosmos) holds a white line sketch of humanoid robots on steps. Bottom-centre circle holds a photograph of Earth from space with a gold and silver car sketch and a robot figure with raised hand overlaid. Green curved double-headed arrows connect Training to Simulation and each circle to the Earth. Right panel: stacked column chart titled NVIDIA Physical AI Revenue ($ Billions). Y axis 0.0 to 20.0 in steps of 4.0, no unit beyond $ Billions in the title. X axis categories FY22, FY23, FY24, FY25, FY26, FY27 (1H). Printed totals: FY22 2.9, FY23 3.3, FY24 5.2, FY25 9.3, FY26 8.2, FY27 (1H) 10.8. Segments are not labelled; about values for FY27 (1H), bottom to top: Auto about 6.4, Healthcare about 0.8, Manufacturing about 1.4, Smart Spaces about 0.8, Telco about 1.1, Robotics about 0.3. Series colours: Auto #76B900, Healthcare #1DBBA4, Manufacturing #0074DF, Smart Spaces #952FC6, Telco #EF9100, Robotics #757575. Legend sits in one row above the plot. FY27 column is highlighted with a lime #B8FF3C outline and a dark-to-grey gradient strip above it, and a rounded callout box with a thin lime leader line points to it.

**Design**

Background: Solid black #000000. Pure black #000000 background. Title white #FFFFFF in NVIDIASans Medium, sans-serif, about 4.9 percent of slide height. Subtitle lines grey #A1A1A1 in NVIDIASans Regular, about 2.0 percent of height. Panel labels white bold NVIDIASans Bold about 1.7 percent. Chart labels white NVIDIASans Regular about 1.5 to 1.7 percent. Accent green #76B900 and lime #B8FF3C. Panels have thin grey #3A3A3A strokes, small corner radius, no shadows.

**Elements and positions** (back to front; x, y, width, height as % of the slide; type, colour and alignment in brackets)

- shape at x 0, y 0, w 100, h 100 [#000000]: Full-slide background rectangle
- shape at x 4.9, y 31, w 44.4, h 54.8 [#000000]: Left diagram panel, rounded rectangle (No fill, 1px stroke #3A3A3A, corner radius about 0.5 percent of width)
- shape at x 50.7, y 31, w 44.4, h 54.8 [#000000]: Right chart panel, rounded rectangle (No fill, 1px stroke #3A3A3A, corner radius about 0.5 percent of width)
- illustration at x 15.4, y 36.7, w 8.8, h 15.6 [#000000]: Training circle: white line sketch of server racks with gold highlights, cropped to a circle (Circular crop with thin white #FFFFFF outline and faint glow)
- illustration at x 29.6, y 36.7, w 8.8, h 15.6 [#000000]: Simulation circle: white line sketch of humanoid robots on steps, cropped to a circle (Circular crop with thin white #FFFFFF outline and faint glow)
- photo at x 22.4, y 59.4, w 9.4, h 16.8 [#1F3A5F]: Model Deployment circle: photograph of Earth from space with a gold and silver car sketch and a robot figure with raised hand (Circular crop, blue Earth with white clouds, dark surround, faint glow)
- line at x 24.4, y 39.1, w 4.9, h 1.3 [#76B900]: Curved double-headed arrow between Training and Simulation circles (Stroke about 2px, arrowheads at both ends)
- line at x 19.3, y 54.2, w 2.8, h 7.8 [#76B900]: Curved double-headed arrow from Training circle to Earth circle (Stroke about 2px, arrowheads at both ends)
- line at x 31.8, y 54.2, w 2.8, h 8.3 [#76B900]: Curved double-headed arrow from Simulation circle to Earth circle (Stroke about 2px, arrowheads at both ends)
- shape at x 57.5, y 58.5, w 32.5, h 13.4 [#757575]: FY22 to FY27 stacked bars, plot area (Bars filled with series colours; FY27 column is the highlighted one)
- shape at x 57.5, y 65.9, w 32.5, h 10.8 [#76B900]: FY27 (1H) Auto segment (Outlined in lime #B8FF3C, about 0.3 percent stroke)
- shape at x 57.5, y 64.5, w 32.5, h 10.8 [#1DBBA4]: FY27 (1H) Healthcare segment
- shape at x 57.5, y 62.1, w 32.5, h 12.6 [#0074DF]: FY27 (1H) Manufacturing segment
- shape at x 57.5, y 60.7, w 32.5, h 12.9 [#952FC6]: FY27 (1H) Smart Spaces segment
- shape at x 57.5, y 58.9, w 32.5, h 13.8 [#EF9100]: FY27 (1H) Telco segment
- shape at x 57.5, y 58.5, w 32.5, h 13.4 [#757575]: FY27 (1H) Robotics segment
- shape at x 87.2, y 34.1, w 2.8, h 24.6 [#757575]: Gradient strip above FY27 column, dark to grey (Image 81 by 399 px, vertical gradient from black at top to grey #757575 at base, sits above the FY27 bar)
- shape at x 76.8, y 42.9, w 9.4, h 13.2 [#000000]: Callout box, rounded rectangle (Black fill, 1.5px stroke #B8FF3C, corner radius about 1.5 percent of width, overlaps the FY26 and FY27 area)
- line at x 81.3, y 56.2, w 5.8, h 6.8 [#B8FF3C]: Leader line from callout box to top-left of FY27 bar (Thin 1px stroke)
- logo at x 92.9, y 94, w 4.6, h 1.8 [#76B900]: NVIDIA logo: green eye mark and white NVIDIA wordmark (Bottom-right corner)
- title at x 18.1, y 10.5, w 63.7, h 4.9 [NVIDIASans Medium, sans-serif, regular width; font size 4.94% of slide height; #FFFFFF; center aligned]: Physical AI Adds a New Workload on AI Factories
- subtitle at x 19.8, y 18.1, w 60.7, h 2 [NVIDIASans Regular, sans-serif; font size 1.98% of slide height; #A1A1A1; center aligned]: Physical AI extends AI into the physical economy – manufacturing, robotics, autonomous vehicles and heavy industry.
- subtitle at x 23, y 20.7, w 54.3, h 2 [NVIDIASans Regular, sans-serif; font size 1.98% of slide height; #A1A1A1; center aligned]: It addresses a $50 trillion market and represents a multi-trillion dollar long-term opportunity for NVIDIA.
- subtitle at x 13.7, y 23.3, w 72.9, h 2 [NVIDIASans Regular, sans-serif; font size 1.98% of slide height; #A1A1A1; center aligned]: Only NVIDIA delivers all three computers necessary to build physical AI – Train (DGX), Simulate (Omniverse + Cosmos), and Run (Jetson Thor).
- text at x 9.4, y 43.5, w 3.6, h 1.7 [NVIDIASans Bold, sans-serif; font size 1.67% of slide height; #FFFFFF; right aligned]: Training
- text at x 10.3, y 45.5, w 1.7, h 1.5 [NVIDIASans Regular, sans-serif; font size 1.48% of slide height; #FFFFFF; right aligned]: DGX
- text at x 41.1, y 43.5, w 5.1, h 1.7 [NVIDIASans Bold, sans-serif; font size 1.67% of slide height; #FFFFFF; left aligned]: Simulation
- text at x 39.5, y 45.5, w 8, h 1.5 [NVIDIASans Regular, sans-serif; font size 1.48% of slide height; #FFFFFF; left aligned]: Omniverse / Cosmos
- text at x 22.9, y 79.5, w 8.4, h 1.7 [NVIDIASans Bold, sans-serif; font size 1.67% of slide height; #FFFFFF; center aligned]: Model Deployment
- text at x 24.7, y 81.4, w 4.8, h 1.5 [NVIDIASans Regular, sans-serif; font size 1.48% of slide height; #FFFFFF; center aligned]: Jetson / IGX
- text at x 62.8, y 34.3, w 20.2, h 2 [NVIDIASans Medium, sans-serif; font size 1.98% of slide height; #FFFFFF; center aligned]: NVIDIA Physical AI Revenue ($ Billions)
- text at x 57.7, y 38.2, w 28, h 1.5 [NVIDIASans Regular, sans-serif; font size 1.48% of slide height; #FFFFFF; left aligned]: Auto / Healthcare / Manufacturing / Smart Spaces / Telco / Robotics (legend, one row, colour squares #76B900, #1DBBA4, #0074DF, #952FC6, #EF9100, #757575 before each label) (Individual labels: Auto at 57.7, Healthcare at 61.0, Manufacturing at 66.8, Smart Spaces at 74.1, Telco at 81.0, Robotics at 84.5 (all y 38.2))
- text at x 53.1, y 42, w 2, h 1.7 [NVIDIASans Regular, sans-serif; font size 1.73% of slide height; #FFFFFF; right aligned]: 20.0
- text at x 53.1, y 48.9, w 2, h 1.7 [NVIDIASans Regular, sans-serif; font size 1.73% of slide height; #FFFFFF; right aligned]: 16.0
- text at x 53.1, y 55.6, w 2, h 1.7 [NVIDIASans Regular, sans-serif; font size 1.73% of slide height; #FFFFFF; right aligned]: 12.0
- text at x 53.6, y 62.3, w 1.4, h 1.7 [NVIDIASans Regular, sans-serif; font size 1.73% of slide height; #FFFFFF; right aligned]: 8.0
- text at x 53.6, y 69.1, w 1.4, h 1.7 [NVIDIASans Regular, sans-serif; font size 1.73% of slide height; #FFFFFF; right aligned]: 4.0
- text at x 53.6, y 75.8, w 1.4, h 1.7 [NVIDIASans Regular, sans-serif; font size 1.73% of slide height; #FFFFFF; right aligned]: 0.0
- text at x 58.2, y 69.6, w 1.4, h 1.7 [NVIDIASans Bold, sans-serif; font size 1.73% of slide height; #FFFFFF; center aligned]: 2.9
- text at x 64.2, y 68.8, w 1.4, h 1.7 [NVIDIASans Bold, sans-serif; font size 1.73% of slide height; #FFFFFF; center aligned]: 3.3
- text at x 70.1, y 65.7, w 1.4, h 1.7 [NVIDIASans Bold, sans-serif; font size 1.73% of slide height; #FFFFFF; center aligned]: 5.2
- text at x 76, y 58.8, w 1.4, h 1.7 [NVIDIASans Bold, sans-serif; font size 1.73% of slide height; #FFFFFF; center aligned]: 9.3
- text at x 82, y 60.6, w 1.4, h 1.7 [NVIDIASans Bold, sans-serif; font size 1.73% of slide height; #FFFFFF; center aligned]: 8.2
- text at x 87.6, y 55.8, w 2.7, h 1.7 [NVIDIASans Bold, sans-serif; font size 1.73% of slide height; #FFFFFF; center aligned]: 10.8
- text at x 90.6, y 55.8, w 1.9, h 1.7 [NVIDIASans Regular, sans-serif; font size 1.73% of slide height; #FFFFFF; left aligned]: (1H)
- text at x 77.9, y 45.5, w 7.1, h 1.7 [NVIDIASans Medium, sans-serif; font size 1.73% of slide height; #FFFFFF; center aligned]: 1H: $10.8B YTD
- text at x 80, y 48.5, w 2.9, h 1.5 [NVIDIASans Regular, sans-serif; font size 1.48% of slide height; #FFFFFF; center aligned]: Already
- text at x 78.6, y 50.4, w 5.7, h 1.7 [NVIDIASans Bold, sans-serif; font size 1.73% of slide height; #B8FF3C; center aligned]: $2.6B above
- text at x 78.7, y 52.4, w 5.6, h 1.5 [NVIDIASans Regular, sans-serif; font size 1.48% of slide height; #FFFFFF; center aligned]: full-year FY26
- text at x 57.8, y 78.1, w 2.3, h 1.7 [NVIDIASans Regular, sans-serif; font size 1.73% of slide height; #FFFFFF; center aligned]: FY22
- text at x 63.7, y 78.1, w 2.3, h 1.7 [NVIDIASans Regular, sans-serif; font size 1.73% of slide height; #FFFFFF; center aligned]: FY23
- text at x 69.6, y 78.1, w 2.3, h 1.7 [NVIDIASans Regular, sans-serif; font size 1.73% of slide height; #FFFFFF; center aligned]: FY24
- text at x 75.6, y 78.1, w 2.3, h 1.7 [NVIDIASans Regular, sans-serif; font size 1.73% of slide height; #FFFFFF; center aligned]: FY25
- text at x 81.5, y 78.1, w 2.3, h 1.7 [NVIDIASans Regular, sans-serif; font size 1.73% of slide height; #FFFFFF; center aligned]: FY26
- text at x 87.5, y 78.1, w 2.3, h 1.7 [NVIDIASans Regular, sans-serif; font size 1.73% of slide height; #FFFFFF; center aligned]: FY27 (File also contains an overlapping text box reading FY271H at x 86.8, w 3.6, same baseline)

**Measured from the file** (exact, read from the PDF rather than estimated; x, y, w, h as % of the slide; font size as % of the slide height)

- fill at x 0.0, y 0.0, w 100.0, h 100.0: #000000 (covers the slide)
- fill at x 57.5, y 65.9, w 32.5, h 10.8: #76B900
- fill at x 57.5, y 64.5, w 32.5, h 10.8: #1DBBA4
- fill at x 57.5, y 62.1, w 32.5, h 12.6: #0074DF
- fill at x 57.5, y 60.7, w 32.5, h 12.9: #952FC6
- fill at x 57.5, y 58.9, w 32.5, h 13.8: #EF9100
- fill at x 57.5, y 58.5, w 32.5, h 13.4: #757575
- image at x 14.6, y 35.4, w 24.6, h 42.9: 1352 by 1328 pixels
- image at x 87.2, y 34.1, w 2.8, h 24.6: 81 by 399 pixels
- image at x 87.2, y 34.1, w 2.8, h 24.6: 81 by 399 pixels
- image at x 81.4, y 56.0, w 5.9, h 7.3: 170 by 119 pixels
- image at x 81.4, y 56.0, w 5.9, h 7.3: 170 by 119 pixels
- text at x 18.1, y 10.5, w 63.7, h 4.9 [NVIDIASans-Medium; font size 4.94; #FFFFFF]: Physical AI Adds a New Workload on AI Factories
- text at x 19.8, y 18.1, w 60.7, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: Physical AI extends AI into the physical economy – manufacturing, robotics, autonomous vehicles and heavy industry.
- text at x 23.0, y 20.7, w 54.3, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: It addresses a $50 trillion market and represents a multi-trillion dollar long-term opportunity for NVIDIA.
- text at x 13.7, y 23.3, w 72.9, h 2.0 [NVIDIASans-Regular; font size 1.98; #A1A1A1]: Only NVIDIA delivers all three computers necessary to build physical AI – Train (DGX), Simulate (Omniverse + Cosmos), and Run (Jetson Thor).
- text at x 62.8, y 34.3, w 20.2, h 2.0 [NVIDIASans-Medium; font size 1.98; #FFFFFF]: NVIDIA Physical AI Revenue ($ Billions)
- text at x 53.1, y 42.0, w 2.0, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: 20.0
- text at x 9.4, y 43.5, w 3.6, h 1.7 [NVIDIASans-Bold; font size 1.67; #FFFFFF]: Training
- text at x 41.1, y 43.5, w 5.1, h 1.7 [NVIDIASans-Bold; font size 1.67; #FFFFFF]: Simulation
- text at x 57.7, y 38.2, w 1.9, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Auto
- text at x 61.0, y 38.2, w 4.3, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Healthcare
- text at x 66.8, y 38.2, w 5.8, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Manufacturing
- text at x 74.1, y 38.2, w 5.4, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Smart Spaces
- text at x 81.0, y 38.2, w 2.1, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Telco
- text at x 84.5, y 38.2, w 3.4, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Robotics
- text at x 77.9, y 45.5, w 7.1, h 1.7 [NVIDIASans-Medium; font size 1.73; #FFFFFF]: 1H: $10.8B YTD
- text at x 53.1, y 48.9, w 2.0, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: 16.0
- text at x 78.6, y 50.4, w 5.7, h 1.7 [NVIDIASans-Bold; font size 1.73; #B8FF3C]: $2.6B above
- text at x 10.3, y 45.5, w 1.7, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: DGX
- text at x 39.5, y 45.5, w 8.0, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Omniverse / Cosmos
- text at x 53.1, y 55.6, w 2.0, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: 12.0
- text at x 80.0, y 48.5, w 2.9, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Already
- text at x 87.6, y 55.8, w 2.7, h 1.7 [NVIDIASans-Bold; font size 1.73; #FFFFFF]: 10.8
- text at x 90.6, y 55.8, w 1.9, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: (1H)
- text at x 76.0, y 58.8, w 1.4, h 1.7 [NVIDIASans-Bold; font size 1.73; #FFFFFF]: 9.3
- text at x 82.0, y 60.6, w 1.4, h 1.7 [NVIDIASans-Bold; font size 1.73; #FFFFFF]: 8.2
- text at x 78.7, y 52.4, w 5.6, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: full-year FY26
- text at x 53.6, y 62.3, w 1.4, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: 8.0
- text at x 70.1, y 65.7, w 1.4, h 1.7 [NVIDIASans-Bold; font size 1.73; #FFFFFF]: 5.2
- text at x 53.6, y 69.1, w 1.4, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: 4.0
- text at x 64.2, y 68.8, w 1.4, h 1.7 [NVIDIASans-Bold; font size 1.73; #FFFFFF]: 3.3
- text at x 58.2, y 69.6, w 1.4, h 1.7 [NVIDIASans-Bold; font size 1.73; #FFFFFF]: 2.9
- text at x 53.6, y 75.8, w 1.4, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: 0.0
- text at x 57.8, y 78.1, w 2.3, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: FY22
- text at x 63.7, y 78.1, w 2.3, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: FY23
- text at x 69.6, y 78.1, w 2.3, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: FY24
- text at x 75.6, y 78.1, w 2.3, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: FY25
- text at x 81.5, y 78.1, w 2.3, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: FY26
- text at x 86.8, y 78.1, w 3.6, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: FY271H
- text at x 87.5, y 78.1, w 2.3, h 1.7 [NVIDIASans-Regular; font size 1.73; #FFFFFF]: FY27
- text at x 22.9, y 79.5, w 8.4, h 1.7 [NVIDIASans-Bold; font size 1.67; #FFFFFF]: Model Deployment
- text at x 24.7, y 81.4, w 4.8, h 1.5 [NVIDIASans-Regular; font size 1.48; #FFFFFF]: Jetson / IGX

### Slide 28

**Review notes**

- Review: The deck reads better as a pre-read than as a projected presentation, and the closing slide 28 is a plain text block that ends weakly.
- What would change the score: I would rebuild slide 28 as a visual summary tied to the three growth pillars, and cut the intro paragraphs above the charts on slides 11 and 15 to one line each.

**Text**

NVIDIA: A Platform for Compounding Growth and Value
NVIDIA is a platform that delivers compounding growth and value to investors.
CUDA programmability underpins the productivity, fungibility, and durability of NVIDIA compute. Our platform addresses every model, workload, and deployment. As a result, NVIDIA is increasing share within an expanding AI market.
NVIDIA is also capturing more value per GW driven by advancements in the GPU and a broadening product portfolio – a reflection of NVIDIA's strategy in becoming a full-stack AI factory platform.
And as the platform broadens from chip to a full-stack AI factory, NVIDIA captures a growing % of customers' IT CapEx.
Together, these three growth pillars will drive an increase in NVIDIA's cash generation and capital return to shareholders.

**Layout**

Single full-bleed black slide with one text column. The title is centred horizontally near the top, about 10.5% from the top edge. Body paragraphs are left-aligned at about 5.2% from the left edge, starting at about 26.6% from the top and stacked down to about 71% with uneven vertical gaps between paragraphs. The lower third is empty. A small NVIDIA logo sits in the bottom-right corner. There is no footer text and no slide number.

**Visuals**

none

**Design**

Background: Solid black #000000 fill covering the whole slide. Solid black background. Title in NVIDIASans Medium, white #FFFFFF, roughly 4.9% of slide height. Body text in NVIDIASans Regular, grey #A1A1A1, roughly 2.7% of slide height. Emphasised phrases are set in NVIDIASans Bold in NVIDIA green #76B900 inline within the grey body lines. No shapes, rules, gradients, shadows or corner radii. Only the logo uses green.

**Elements and positions** (back to front; x, y, width, height as % of the slide; type, colour and alignment in brackets)

- shape at x 0, y 0, w 100, h 100 [#000000]: Full-slide background rectangle (Solid fill, no outline, covers the entire slide.)
- title at x 13.6, y 10.5, w 73.4, h 4.9 [NVIDIASans Medium; font size 4.94% of slide height; #FFFFFF; center aligned]: NVIDIA: A Platform for Compounding Growth and Value
- body at x 5.2, y 26.6, w 55.7, h 2.7 [NVIDIASans Regular; font size 2.72% of slide height; #A1A1A1; left aligned]: NVIDIA is a platform that delivers compounding growth and value to investors.
- body at x 5.2, y 35.7, w 86, h 2.7 [NVIDIASans Regular; font size 2.72% of slide height; #A1A1A1; left aligned]: CUDA programmability underpins the productivity, fungibility, and durability of NVIDIA compute. Our platform addresses
- body at x 5.2, y 39.6, w 37.1, h 2.7 [NVIDIASans Regular; font size 2.72% of slide height; #A1A1A1; left aligned]: every model, workload, and deployment. As a result,
- body at x 42.3, y 39.6, w 42.8, h 2.7 [NVIDIASans Bold; font size 2.72% of slide height; #76B900; left aligned]: NVIDIA is increasing share within an expanding AI market.
- body at x 5.2, y 48.7, w 31.8, h 2.7 [NVIDIASans Bold; font size 2.72% of slide height; #76B900; left aligned]: NVIDIA is also capturing more value per GW
- body at x 37.5, y 48.7, w 51.7, h 2.7 [NVIDIASans Regular; font size 2.72% of slide height; #A1A1A1; left aligned]: driven by advancements in the GPU and a broadening product portfolio –
- body at x 5.3, y 52.6, w 54.8, h 2.7 [NVIDIASans Regular; font size 2.72% of slide height; #A1A1A1; left aligned]: a reflection of NVIDIA's strategy in becoming a full-stack AI factory platform.
- body at x 5.3, y 61.7, w 47, h 2.7 [NVIDIASans Regular; font size 2.72% of slide height; #A1A1A1; left aligned]: And as the platform broadens from chip to a full-stack AI factory,
- body at x 52.3, y 61.7, w 38.9, h 2.7 [NVIDIASans Bold; font size 2.72% of slide height; #76B900; left aligned]: NVIDIA captures a growing % of customers' IT CapEx.
- body at x 5.3, y 70.8, w 85.8, h 2.7 [NVIDIASans Regular; font size 2.72% of slide height; #A1A1A1; left aligned]: Together, these three growth pillars will drive an increase in NVIDIA's cash generation and capital return to shareholders.
- logo at x 92.9, y 94.2, w 4.6, h 1.6 [#76B900]: NVIDIA logo: green eye mark with NVIDIA wordmark in white-on-black, bottom right corner (Estimated box; no logo fill colour read from file. Wordmark appears in green as well.)

**Measured from the file** (exact, read from the PDF rather than estimated; x, y, w, h as % of the slide; font size as % of the slide height)

- fill at x 0.0, y 0.0, w 100.0, h 100.0: #000000 (covers the slide)
- text at x 13.6, y 10.5, w 73.4, h 4.9 [NVIDIASans-Medium; font size 4.94; #FFFFFF]: NVIDIA: A Platform for Compounding Growth and Value
- text at x 5.2, y 26.6, w 55.7, h 2.7 [NVIDIASans-Regular; font size 2.72; #A1A1A1]: NVIDIA is a platform that delivers compounding growth and value to investors.
- text at x 5.2, y 35.7, w 86.0, h 2.7 [NVIDIASans-Regular; font size 2.72; #A1A1A1]: CUDA programmability underpins the productivity, fungibility, and durability of NVIDIA compute. Our platform addresses
- text at x 5.2, y 39.6, w 37.1, h 2.7 [NVIDIASans-Regular; font size 2.72; #A1A1A1]: every model, workload, and deployment. As a result,
- text at x 42.3, y 39.6, w 42.8, h 2.7 [NVIDIASans-Bold; font size 2.72; #76B900]: NVIDIA is increasing share within an expanding AI market.
- text at x 5.2, y 48.7, w 31.8, h 2.7 [NVIDIASans-Bold; font size 2.72; #76B900]: NVIDIA is also capturing more value per GW
- text at x 37.5, y 48.7, w 51.7, h 2.7 [NVIDIASans-Regular; font size 2.72; #A1A1A1]: driven by advancements in the GPU and a broadening product portfolio –
- text at x 5.3, y 52.6, w 54.8, h 2.7 [NVIDIASans-Regular; font size 2.72; #A1A1A1]: a reflection of NVIDIA's strategy in becoming a full-stack AI factory platform.
- text at x 5.3, y 61.7, w 47.0, h 2.7 [NVIDIASans-Regular; font size 2.72; #A1A1A1]: And as the platform broadens from chip to a full-stack AI factory,
- text at x 52.3, y 61.7, w 38.9, h 2.7 [NVIDIASans-Bold; font size 2.72; #76B900]: NVIDIA captures a growing % of customers' IT CapEx.
- text at x 5.3, y 70.8, w 85.8, h 2.7 [NVIDIASans-Regular; font size 2.72; #A1A1A1]: Together, these three growth pillars will drive an increase in NVIDIA's cash generation and capital return to shareholders.

### Slide 29

**Text**

NVIDIA

**Layout**

Full-bleed photographic background with the NVIDIA logo lockup placed in the lower-left area; no title, body text or slide number. The 3D green logo mark sits on the right third and bleeds off the right and top edges.

**Visuals**

Photograph-style 3D render: a large glossy green NVIDIA eye-shaped logo mark, extruded and seen at an oblique angle, floating over a dark field of faceted, triangular black tiles that recede toward the upper left. Lighting is directional from the upper right, producing bright specular highlights on the green edges and a grey-blue sheen on the upper-right tiles. Composition is landscape, the logo cropped by the right and top edges. No overlay or tint beyond natural darkness. Small NVIDIA logo lockup at lower left: a green eye mark followed by the white wordmark NVIDIA.

**Design**

Background: Near-black #0A0C0E faceted 3D triangular tile surface, photographed/rendered behind all content. Dark, near-black background with faceted triangular 3D texture. Accent colour bright NVIDIA green #76B900 for the logo mark. Wordmark in white #FFFFFF, bold geometric sans-serif capitals. No titles or body text. No rounded corners, borders or shadows beyond the render's own soft shading.

**Elements and positions** (back to front; x, y, width, height as % of the slide; type, colour and alignment in brackets)

- photo at x 0, y 0, w 100, h 100 [#0A0C0E]: Dark faceted triangular 3D tile surface with a large glossy green 3D NVIDIA eye-logo mark at right, bleeding off the right and top edges (Full-bleed background image; the green logo mark occupies roughly x 54-100, y 13-86; highlights on green edges; mark colour #76B900)
- logo at x 5, y 86, w 15, h 5 [Bold geometric sans-serif, all capitals; #FFFFFF; left aligned]: NVIDIA logo lockup: green eye mark with white wordmark NVIDIA (Green eye mark #76B900 at about x 5-9.3, y 86-91.5; white wordmark about x 10-20, y 87-90.8; sits over dark tiles with no backing box)
