Nvidia’s latest 10-Q, filed on 26 August, discloses $169.5 billion of ecosystem commitments and guarantee caps not yet paid out, by our sum: $25 billion of equity investment commitments and $36 billion of AI cloud service commitments at 26 July, plus $108.5 billion of maximum guarantee exposure including $105 billion signed in August. That is about 3.0 times the $56.6 billion of cash, cash equivalents and marketable debt securities held on 26 July, the latest balance-sheet date, by our arithmetic.
Venture capitalists, startup founders and semiconductor executives expect Nvidia to pull back on its dealmaking pace after the chief executive’s resignation, The Information reported at 19:30 UTC on 12 November. Nvidia’s new chair said on 12 November that the board expects customer commitments to continue without interruption.
Business Pill · Concentration Explained
An evergreen Business Engineer explainer on concentration, where a few holdings account for most of the value; in Nvidia’s 10-Q, guarantee caps are 64% of the $169.5 billion of open commitments, by our arithmetic. The video teaches the general idea only and does not discuss Nvidia.
The key insight: As we read the 10-Q, most of what Nvidia has pledged to its ecosystem is not yet cash out the door: $169.5 billion of commitments and guarantee caps sit beside $99 billion already invested, and the largest part, $108.5 billion, is contingent credit support rather than equity.
The Information said Nvidia has emerged as “one of the most active backers of AI startups, small cloud providers and frontier AI labs” over the last three years, and that its “frenetic dealmaking pace” “totaled $140 billion, including massive credit guarantees, in just two months this year”.
It added that “the most vulnerable could be the companies in the middle of funding rounds in which Nvidia is expected to invest”. We read the headline and the two opening paragraphs shown in The Information’s public feed; the rest of the article is paywalled.
Nvidia’s disclosed equity investment commitments and AI cloud service commitments at 26 July 2026 and maximum guarantee exposure including the August 2026 guarantees (10-Q filed 26 August), against first-half free cash flow and cash plus marketable debt securities. The $169.5 billion sum, the $70.0 billion free cash flow and the ratios are our arithmetic.
What Nvidia Has Committed, Per Its Filings
The 10-Q for the quarter ended 26 July says Nvidia’s ecosystem investments “include equity investments of $99 billion and equity investment commitments of $25 billion as of July 26, 2026”. A commitments table schedules $18 billion of the $25 billion for the rest of fiscal 2027, and a note says the equity commitments, to “AI model makers, infrastructure financiers, and other private companies”, are “subject to certain contingencies”.
Under a model introduced that quarter, AI cloud partners buy Nvidia’s data center products and Nvidia commits to cloud service agreements. Those commitments, “typically six years in duration”, “totaled $36 billion as of July 26, 2026”, and the partners “can unilaterally stop providing to us and sell to third-party customers at more advantageous rates”, the 10-Q says.
The largest item is guarantees: maximum gross exposure of $108.5 billion, of which $105 billion is the cap on credit support for an affiliate of OpenAI at SB Energy’s Ohio campus and $3.5 billion covers other AI cloud partners. The filing says the guarantees are “limited to defined portions of lease and power payments and not the full cost of the site or all of the tenant’s obligations”.
How Much That Is, by Our Arithmetic
The three items add up to $169.5 billion, of which guarantee caps are 64%. Against first-half free cash flow of about $70.0 billion (operating cash flow of $74.4 billion minus $4.4 billion of capital purchases), that is about 2.4 times, by our arithmetic.
Counting the $42.8 billion of marketable equity securities Nvidia also held, liquid holdings were $99.4 billion, and the $169.5 billion is about 1.7 times that. The $99 billion already invested plus the $25 billion of equity commitments comes to $124 billion.
In the 10-K for the year to 25 January 2026, investment commitments were $11.4 billion and guarantee exposure $3.5 billion, $14.9 billion in all, by our sum, against $169.5 billion in the August 10-Q, about 11 times as much; the AI cloudmodel did not yet exist. Equity commitments rose about 2.2 times between the two balance-sheet dates. In the same half, Nvidia spent $42.4 billion buying equity securities, and its privately held stakes rose from $22.3 billion to $47.9 billion at carrying value.
What the 8-K and the Interim CEO Said
Nvidia’s 8-K says the board accepted the resignation on 11 November and appointed its finance chief, Colette Kress, “to serve as interim Chief Executive Officer, while the Board conducts a search for a new Chief Executive Officer”. The board also named Harvey C. Jones as chair.
In the release, Jones said: “The Board has full confidence in Colette and the leadership team, and we expect the company’s operations, customer commitments and product roadmap to continue without interruption.” Kress said: “We will continue advancing the work that has made NVIDIA a leader in accelerated computing and AI, and execute on the opportunities ahead.”
Neither document mentions the investment programme or the guarantees. The Information has also reported that Kress, as finance chief, pushed back on a proposed $250 billion OpenAI backstop, as covered in our piece on the OpenAI backstop.
Where the Filings Leave Room to Decide
The filings set out three points where the amounts can change: the equity commitments are subject to contingencies; the AI cloud commitments shrink as capacity is used by others; and Nvidia holds an option, “exercisable in our sole discretion”, on credit support for about 3.8 more gigawatts in Ohio.
As we read the filings, the amounts can move at three points the 10-Q names: contingencies on the $25 billion of equity commitments, AI clouds selling committed capacity to others, and an Ohio option Nvidia may exercise at its sole discretion.
As we read the 8-K and the release together, the board has said customer commitments continue; neither document addresses the investment commitments or guarantees, which is where The Information reported investors expect a pull-back.
Nvidia 10-Q for the quarter ended 26 July 2026
“These arrangements require substantial payments over many years. If future demand, our needs or plans differ from our expectations, we may be unable to reduce these commitments.”
Three Implications
FOR STARTUPS MID-ROUND The Information reported that companies in the middle of funding rounds in which Nvidia is expected to invest could be the most vulnerable; the 10-Q schedules $18 billion of the $25 billion of equity commitments for the rest of fiscal 2027.
FOR AI CLOUD PARTNERS The 10-Q puts Nvidia’s AI cloud service commitments at $36 billion, typically six years long, shrinking as capacity is used by others.
FOR NVIDIA INVESTORS Nvidia reports third-quarter results on 17 November; the next 10-Q will update the commitments, guarantees and cash figures in this piece.
The framework argues: “It is a sovereign-scale financing problem dressed in tech-company clothing — and the financial architecture is starting to invert.”
As we read it, a chip supplier carrying $169.5 billion of equity commitments, cloud commitments and lease guarantees for its own customers is one form of that inversion. The framework predates the report and does not discuss Nvidia’s leadership change.
Nvidia’s 10-K (25 January 2026): $11.4 billion of investment commitments plus $3.5 billion of guarantee exposure. 10-Q filed 26 August 2026: $25 billion of equity commitments and $36 billion of AI cloud service commitments at 26 July, plus $108.5 billion of guarantee exposure including August. Totals and the 11x ratio are our arithmetic.
Update, 13 November 2026: Nscale and Firmus, two Nvidia-backed cloud builders. The Information reported on 12 November (19:30 UTC) that deals at Nscale and Firmus “hinge on Nvidia extending” its investment sprint, and that Nvidia’s role as one of the most active backers of AI startups, small cloud providers and frontier labs “now looks at risk” after Jensen Huang’s resignation as chief executive. We could read only the headline and summary; the article is paywalled. What the primaries show. Nscale’s S-1, filed 18 September 2026, says Nvidia holds more than 5% of its share capital since the October 2025 Series B; in the Series C, Nvidia’s $300.0 million of pre-Series C SAFEs converted into 16,346,160 preferred shares and Nvidia bought a further 23,410,920 preferred shares for $477.4 million, $777.4 million in that round by our arithmetic. Firmus’s 8 September release puts its contracted capacity at more than 900 MW, with OpenAI as an anchor customer; its portfolio spans seven AI factories in four countries, two operational and five under development, and the factories under the partnership will use Nvidia Vera Rubin NVL72 systems. Bloomberg described Firmus as Nvidia-backed in October, after its planned IPO of about $5 billion collapsed, and The Information reported on 20 October that Nvidia was considering another major investment as Firmus sought at least $2 billion privately. We found no Firmus or Nvidia release stating the size of Nvidia’s Firmus stake or any new terms.
Update, 13 November 2026: what The Wall Street Journal reported about an August finance meeting. The Wall Street Journal reported that Jensen Huang, then Nvidia’s chief executive, met the company’s senior finance team in August over a “ballooning problem”: “Nvidia was in talks to provide billions of dollars of guarantees to companies that couldn’t afford its chips”. The Journal said the deals included financing offered to “upstart cloud providers and data-center developers” in the hopes of turning them into long-term customers, and that Huang “was now starting to question whether it had gone too far, according to people familiar with the matter”. Per the Journal, Huang described major cloud providers “making plans to use alternative chips and lessen their reliance on the firm”, and said the “vast supply of chip wafers and memory Nvidia had committed to buying” was coming like an “avalanche”. We could not open the full article; these lines are from the Journal’s text as quoted in an X post by investor Michael Burry at 04:34 UTC on 13 November. Our context, from the 10-Q filed 26 August: Nvidia’s supply and capacity commitments, “primarily memory and manufacturing facilities”, rose from $119 billion the previous quarter to $279 billion at 26 July, 2.3 times as much by our arithmetic, with $92 billion due in the rest of fiscal 2027. That $279 billion sits outside the $169.5 billion of investment, cloud and guarantee commitments on this page, which includes $108.5 billion of maximum guarantee exposure; together they come to $448.5 billion, by our arithmetic.
Update, 12 November, 22:00 UTC: Axios flags biotech and energy deals. Axios published three headlines at about 21:42 UTC: “Huang’s ouster threatens Nvidia’s breakneck dealmaking pace”, “Nvidia’s biotech dealmaking faces a new test” and “Nvidia turmoil threatens to roil data center energy deals”. We could not access the articles, so we do not report their figures. Nvidia’s 13F for 30 June 2026 lists one biotech stake, 833,325 shares of Generate Biomedicines worth $14.1 million, about 0.02% of the $63.4 billion of listed stakes, by our arithmetic. Its 13F for 30 September 2025 listed 7.7 million Recursion Pharmaceuticals shares worth $37.6 million; that row is absent from the filings for 31 December 2025 onward. None of the four latest 13F filings lists a utility or power producer; Nvidia’s energy-linked exposure in its filings sits in the guarantee caps above, mainly the $105 billion tied to lease and power payments at SB Energy’s Ohio campus.
What Is Not Established
We have not read The Information’s full article, and it is not established from what we read which funding rounds or companies it refers to, or how its $140 billion figure is built. The 10-Q balances are as of 26 July, with the guarantee total updated for August, and predate the leadership change; the 10-Q excerpts do not say whether the Anthropic purchase is part of the $25 billion.
Neither the 8-K nor the release says whether Nvidia’s investment policy will change. Nvidia was not contacted for this piece. Nvidia reports third-quarter results on 17 November, when the next 10-Q will update these figures.
This Week In The AI Supercycle
The Great Convergence of Capital, Power, and Geopolitics
This week’s issue: for two years the AI build-out was priced as a supply story, and last week as a credit story. This week it was priced as a permission story, and the answer arrived as a cost, not a ban.
The Information reported that investors and founders expect Nvidia to slow its dealmaking after the chief executive’s exit. Nvidia’s August 10-Q shows $169.5 billion of equity commitments, AI cloud service commitments and guarantee caps, about 3.0 times its 26 July cash and marketable debt securities, by our arithmetic, against the chair’s statement that the board expects customer commitments to continue without interruption.
95,000+ executives read Business Engineer for the AI strategy frameworks cited by ChatGPT, Claude, and Perplexity.
A note on sourcing. The Information’s article is paywalled: we read its headline, bylines and time, and the two opening paragraphs shown in its public feed, and quote only those. Commitments, guarantees, cash flow and balance-sheet figures come from Nvidia’s 10-Q for the quarter ended 26 July 2026 and its 10-K for the year to 25 January 2026; the appointments and statements from the 8-K and release of 12 November. Sums, ratios and free cash flow are our arithmetic. Nvidia was not contacted for this piece.
Gennaro Cuofano is a CRO and tech executive who has worked in AI since late 2015, starting with bringing NLP, natural-language generation, voice and chatbot products to the marketing industry. His background is in law and finance: he holds a Master's degree in Law and an International MBA with an emphasis on corporate finance (LUISS Business School and the University of San Diego, 2012), and worked as a financial analyst at a real-estate investment firm in San Diego and as an assistant controller. His work focuses on business model strategy, business engineering and, more broadly, structural analysis: how companies make money, read from their own filings. He created FourWeekMBA and leads research there and at The Business Engineer, his newsletter on AI and business strategy, with over 95,000 subscribers and more than 1,000 published analyses. His writing has also appeared on Entrepreneur (for example, How to Digitize a Family Business in Europe) and HackerNoon. Find him on LinkedIn and Substack. See how we source.
