Nvidia has committed $279 billion to suppliers, and its 10-Q for the quarter to 26 July schedules $267 billion of it, 95.7% by our arithmetic, for fiscal 2027 to 2029. Its 2026 annual review letter says Nvidia has “visibility into more than $1 trillion in cumulative Blackwell and Rubin revenue from the start of 2025 through 2027”. Whether it can deliver that roadmap without its founder is what SemiAnalysis argued about on 13 November.
On an emergency episode recorded with ChinaTalk after Nvidia’s chief executive resigned, SemiAnalysis’s Jordan Nanos said “in the short term Nvidia is fine” and called the company “still Nvidia’s to lose”. SemiAnalysis’s Dylan Patel listed where execution could slip: the Rubin yield ramp, Feynman’s timing, the Rubin Ultra specification, software support and supply allocation. We checked both views against Nvidia’s filings and tested the episode’s Apple comparison: Apple returned 31.6 times its money after its 2011 handover, against 8.4 times for an S&P 500 index fund, by our arithmetic.
Business Pill · Forecast vs Commitment Explained
An evergreen Business Engineer explainer on how a date said out loud becomes a commitment. It teaches the concept only and does not discuss Nvidia, its roadmap or the podcast in this piece.
The key insight: As we read the filings against the episode, the argument is about delivery, not demand: Nvidia’s 10-Q schedules $267 billion of supply commitments for fiscal 2027 to 2029, by our arithmetic, the years in which Vera Rubin ships and Rubin Ultra systems are due, so both SemiAnalysis views are tested on the same schedule.
Nanos said Nvidia now has to deliver on commitments to customers with the supply chain it already has. “I think they know their marching orders.” The episode, uploaded to SemiAnalysis’s YouTubechannel on 13 November, runs 51 minutes; the chapter on Nvidia without its founder starts at 31:10.
Nvidia’s own documents set out the roadmap Nanos referred to. Its 10-Q says “Our next-generation Data Center architecture, Vera Rubin, began production shipments in the third quarter of fiscal year 2027”, and that Nvidia is “currently experiencing certain supply constraints”. The GTC press release in March said “Vera Rubin-based products will be available from partners starting the second half of this year”.
For the next step, Nvidia’s GTC 2025 blog said: “And due for the second half of 2027: systems built on Rubin Ultra.” Its technical blog of 16 March 2026 says Vera Rubin Ultra NVL576 “will combine eight separate MGX NVL racks”. The GTC 2026 keynote blog says “Looking beyond Vera Rubin, NVIDIA’s next major architecture is Feynman”; none of the Nvidia documents we read gives Feynman a year. Our piece on SemiAnalysis’s Rubin preview has the benchmark side.
Nvidia’s supply and capacity commitments by fiscal year as of 26 July 2026, from Note 10 of its 10-Q: $92 billion for the rest of fiscal 2027, $87 billion for fiscal 2028, $88 billion for fiscal 2029, then $6 billion, $5 billion and $1 billion, $279 billion in all. The first three add up to $267 billion, 95.7% of the total, by our arithmetic. Labels mark the roadmap windows Nvidia has given: Vera Rubin production shipments began in the third quarter of fiscal 2027 (10-Q), and systems built on Rubin Ultra are due in the second half of 2027 (GTC 2025 blog). Nvidia’s fiscal year ends in late January.
Where Patel Sees Execution Slipping
Asked on what dimensions Nvidia would start to falter, Patel began with manufacturing. If Rubin hits manufacturing issues without the founder there, Patel said, “things get solved slower, things get solved by committees, things don’t get escalated up”.
Patel’s list ran from a worse Rubin yield curve to Feynman slipping or being less aggressive, a slower Rubin Ultra and a Rubin Ultra specification that might take longer or not get done, which would lower the volumes Nvidia can ship, Patel said. On software, Patel said work to run models and kernels from DeepSeek and Moonshot efficiently on Nvidia hardware would take longer than before.
Patel’s example of a founder-driven product call was Volta. “Volta didn’t have Tensor until 3 months before tape out”, Patel said, describing a late design change the founder decided. On supply, Patel pointed to “the supply chain giving them 70% of their supply instead of, you know, much less”, and set out the downside case: “it could be like in 2029 the best chips aren’t Nvidia’s and they don’t have the majority of the supply chain”.
What Nvidia Has Already Committed
The supply side Patel described shows up in Nvidia’s commitments. The 10-Q says Nvidia increased supply commitments from $119 billion the previous quarter to $279 billion as of 26 July 2026, a jump of $160 billion in one quarter, by our arithmetic. The commitments are “primarily memory and manufacturing facilities”, the filing says.
By fiscal year, the filing schedules $92 billion for the rest of fiscal 2027, $87 billion for fiscal 2028 and $88 billion for fiscal 2029, then $6 billion, $5 billion and $1 billion. The first three add up to $267 billion, by our arithmetic. Nvidia’s fiscal year ends in late January, so fiscal 2029 runs to early 2029. Some agreements “may be cancelable, rescheduled, or adjustable for our businessneeds prior to placing firm orders”, the 10-Q says, and changes “may result in additional costs”. Our piece on Nvidia’s deal pledges covers the customer side.
The Retention Test
Nanos put the bigger test on people. “Jensen has been incredibly good at retaining talent”, Nanos said, adding that few executives leave and naming Rene Haas, now Arm’s chief executive, as the main exception. “The VCs are going to come in and offer people billions of dollars to start their own chip companies, their own Neolabs.” Our piece on venture investors circling Nvidia’s Israel teams follows that thread.
Nanos set a one-year marker: “if a year from now we’ve seen like a bunch of departures from Nvidia, I’m going to be really concerned about their future”. Nanos added “I think there’s a very very high chance that they’re still doing just fine 2 3 4 years from now”. Our piece on Nvidia’s executive bench sets out the five executive-officer seats and their years at the company.
The Form 8-K of 12 November says the board appointed Colette M. Kress, Executive Vice President and Chief Financial Officer since 2013, as interim chief executive “while the Board conducts a search for a new Chief Executive Officer”, and that Kress keeps the finance role during that time. In the episode’s closing thoughts, one panelist framed the tension: “Sales wants to do deals. Technology wants to make big bets.
Finance needs to figure out how to pay for it. Finance is in charge now.” The same speaker expected Nvidia to be “very profitable for a little while”. Our guide to the CEO search covers the names reported by news outlets, and the chain began with the after-hours drop.
Total return since Apple named its chief operating officer chief executive on 24 August 2011, from Yahoo Finance adjusted closes, which add back dividends, at month-end on a log scale, 24 August 2011 = 100, to the 12 November 2026 close. Apple returned 31.6 times and the S&P 500 index fund 8.4 times, by our arithmetic. Past returns are not a guide to Nvidia.
The Apple Comparison, Checked
The panel answered people saying Nvidia was “at zero” with Apple: “you’d return a ton of money with Apple stock since that time”. On 24 August 2011, Apple’s board named Tim Cook, previously Apple’s Chief Operating Officer, as chief executive, citing “Tim’s 13 years of service to Apple”.
We checked it. Using Yahoo Finance adjusted closes, which add back dividends, Apple stock bought at the 24 August 2011 close was worth 31.6 times as much at the 12 November 2026 close, a 25.5% yearly return, by our arithmetic. The SPDR S&P 500 ETF Trust, an index fund, returned 8.4 times, or 15.0% a year, by our arithmetic. The panel also gave the counterpoint: Nvidia’s founder was “negotiating every supply chain contract and Steve Jobs is not doing that”.
The Structural Read
As we read Patel’s list, most of the risks are about timing, yield and supply allocation, which are the items a commitment schedule makes visible: the 10-Q says some agreements may be cancelable, rescheduled or adjustable before firm orders, with possible additional costs.
As we read Nanos’s case, the test runs on people rather than parts: the 8-K puts the finance chief in the interim role while the board searches, and the one-year marker Nanos set is the window in which most of the $267 billion is scheduled.
NVIDIA Corporation, Form 10-Q for the quarter ended July 26, 2026
“These supply commitments are for our data center infrastructure systems, primarily memory and manufacturing facilities, to produce our products for long-term demand across current and future product architectures.”
Three Implications
FOR NVIDIA’S CUSTOMERS Vera Rubin began production shipments in the third quarter of fiscal 2027, the 10-Q says, and Nvidia’s GTC 2025 blog put systems built on Rubin Ultra in the second half of 2027.
FOR SUPPLIERS The 10-Q schedules $92 billion of supply and capacity commitments for the rest of fiscal 2027, $87 billion for fiscal 2028 and $88 billion for fiscal 2029.
FOR INVESTORS The 8-K names an interim chief executive while the board searches; Nvidia’s next 10-Q will update the commitment schedule.
The framework describes its subject as “How the money flows, where it cracks, what it buys, who can absorb it, and what happens when revenue can’t catch up.”
As we read it, Nvidia’s $279 billion of supply commitments is the supplier end of that flow, and the execution risks SemiAnalysis listed sit where it could crack. The framework does not discuss Nvidia’s leadership change.
What Is Not Established
The episode’s views are opinions, and its captions are YouTube’s automatic ones without speaker labels; we attribute lines from the episode description and the order of the conversation, and we quote the captions as written. The 2029 scenario, the 70% supply figure and the Volta account are Patel’s, and we could not check them against a filing.
The supply schedule is as of 26 July 2026 and may have changed; the filing says some agreements can be adjusted. Apple’s return is history, not a guide to Nvidia. We leave out the episode’s personal and speculative segments under our coverage rules, and we did not contact Nvidia or SemiAnalysis for this piece.
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.
SemiAnalysis split on Nvidia after its founder: Nanos said the roadmap carries the short term and retention is the test, while Patel listed the places execution could slip, from Rubin yields to Feynman’s timing and supply allocation. Nvidia’s 10-Q schedules $267 billion of supply commitments, by our arithmetic, for fiscal 2027 to 2029, the years in which Vera Rubin ships and Rubin Ultra systems are due. The next filing to watch is Nvidia’s 10-Q for the quarter to late October, with an updated schedule of those commitments.
A note on sourcing. We reviewed SemiAnalysis’s Ep. 044, an emergency episode with ChinaTalk uploaded on YouTube on 13 November 2026, through its description and YouTube’s automatic captions, which carry no speaker labels; quotations are copied from those captions, and speakers are identified from the description and the flow of the conversation. We read, on 13 November, Nvidia’s 10-Q for the quarter to 26 July 2026 and its Form 8-K of 12 November on SEC EDGAR, its 2026 Annual Review, its GTC 2026 Vera Rubin release, technical blog and keynote blog, its GTC 2025 blog, and Apple’s release of 24 August 2011.
Price data are Yahoo Finance adjusted closes for Apple and the SPDR S&P 500 ETF Trust. The $267 billion and 95.7% figures, the $160 billion increase, the 31.6 and 8.4 multiples and the yearly returns are our arithmetic. Under our Nvidia coverage rules we leave out the episode’s personal material and its speculation about individuals. We did not contact Nvidia or SemiAnalysis 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.
