NVIDIA AI compute infrastructure financing platforms

The NVIDIA AI compute infrastructure financing platforms are a set of proposed, independently run financing vehicles that NVIDIA announced on August 10, 2026, together with Apollo, BlackRock, Blackstone, Brookfield, Goldman Sachs and KKR. Under memorandums of understanding, the six financial institutions and NVIDIA said they would establish "independent compute financing platforms" intended to mobilize more than $500 billion of third-party capital over time for the buildout of AI infrastructure across NVIDIA's customer base, including frontier AI labs, enterprises and AI clouds. NVIDIA's stated aim is to have its compute treated as an investable infrastructure asset, so that institutional debt and insurance capital can underwrite GPU-based "AI factories" (data centers built around NVIDIA systems) the way lenders underwrite power plants, aircraft or real estate.^[1][2] The announcement was made through a joint press release, a live CNBC panel featuring Jensen Huang and executives from all six firms, and a long post by Huang on X that NVIDIA republished as a blog the next day.^[1][4][5][3]

The headline figure is a target, not committed money. NVIDIA's release said the partners would mobilize the capital "over time" and that the partnerships "remain subject to execution of the final agreements"; Huang's post added that the $500 billion is aggregate third-party capital the platforms are "designed to mobilize over time" and "not NVIDIA revenue, a single fund or a commitment to a single customer."^[1][3] As of September 4, 2026, NVIDIA had announced no definitive agreement or first transaction; its second-quarter earnings release on August 26 still described the platforms as "subject to definitive agreements," and a September 3 IFR commentary said nothing published since the announcement had filled in the details.^[31][45][43] The plan drew immediate praise from the participating executives and from several sell-side analysts, and immediate criticism from short seller Michael Burry, bond investor Jeffrey Gundlach and others who questioned whether fast-depreciating chips can serve as collateral for long-dated debt and whether the structure genuinely answers the "circular financing" critique of NVIDIA's ecosystem deals.^[18][22][9] The debate sits inside the wider argument over AI-infrastructure debt described at AI bubble.

Announcement

The Financial Times reported the plan on the morning of August 10, 2026, citing six unnamed people familiar with the matter; NVIDIA confirmed it in a press release the same day.^[7][1] The release, datelined Santa Clara and New York, said NVIDIA had signed memorandums of understanding with the six institutions "to establish the first compute financing platforms of their kind at global scale," and that NVIDIA would work with them "to create dedicated pools of capital at significant scale at attractive rates for NVIDIA customers." It described NVIDIA compute as "an investable asset" that "provides the lowest token cost, highest revenue and longest life along with a rich ecosystem of offtakers built upon NVIDIA's CUDA platform."^[1] Apollo and Blackstone republished the same joint release on their own newsrooms the same day.^[11][12]

Huang's quote in the release framed the initiative as a corporate milestone: "We began by building chips; today, we are helping create a new class of productive, investable infrastructure: AI factories." He added that "In AI, compute is revenue," and that NVIDIA compute is "broadly adopted, flexible across models and workloads, fungible and transferable across customers and operators, and continuously improved through CUDA software." The platforms, he said, "will help customers access scarce compute at scale and build the DSX AI factories that will power every industry and country in the age of AI."^[1]

That afternoon Huang appeared on CNBC's "Closing Bell: Overtime" with Goldman Sachs CEO David Solomon, Blackstone president Jon Gray, Apollo president Jim Zelter, Brookfield CEO Bruce Flatt and KKR's global head of digital infrastructure Waldemar Szlezak, with BlackRock CEO Larry Fink joining remotely. Huang told anchor Becky Quick that "this is really the first time that technology chips have become an investable asset class," and that the systems "are revenue-generating assets. Now, they're productive, they're long-lived, they're fungible, they're flexible."^[4][5] He put the cost of building AI capacity at "something like $50, $60 billion" per gigawatt, and said all of the money would be "third-party, independent, long-term capital."^[5] Solomon said that "Jensen came, approached us with the idea," and Bloomberg reported that Huang said he had approached only those six firms and that none turned him down.^[5][6]

At 21:56 UTC on August 10, Huang posted a long-form article on X titled "NVIDIA AI Factory Compute Is Becoming an Investable Asset Class."^[2] NVIDIA published the same text on its corporate blog under Huang's byline, dated August 11, 2026.^[3] The post is the fullest public description of the scheme's logic and of NVIDIA's own role in it, and most subsequent press coverage quotes it.

Partners and stated roles

The release attributes one quote to each partner. Five of the six (Apollo, BlackRock, Blackstone, Brookfield and KKR) are asset managers, and Goldman Sachs is the only bank in the group; KKR's release boilerplate also describes insurance subsidiaries under Global Atlantic.^[1][37] The boilerplate in the release gives Apollo's assets under management as approximately $1.05 trillion as of June 30, 2026, Blackstone's as over $1.3 trillion, and Brookfield's as more than $1 trillion.^[1]

Scroll sideways for more →

Bloomberg reported that Goldman Sachs, as the only bank, was positioning itself to be lead bookrunner on the public debt deals and would also earn returns on debt distributed through its asset-management arm.^[6] Reuters reported on August 14 that Goldman was already in talks with potential investors, that US insurers, money managers and banks were expected to form the core investor base, that asset managers planned to retain a sizable share of the financing, and that Goldman could provide junior capital and private credit through its asset-management business while its investment bank placed debt into private credit funds and, eventually, public debt markets.^[20][21] Reuters also noted Goldman's history with NVIDIA, including a role as lead underwriter on NVIDIA's $25 billion bond sale in June 2026 and as exclusive adviser on the 2019 Mellanox acquisition.^[20]

How the platforms are meant to work

NVIDIA has described the mechanism only in outline. The release says the capital providers will "independently underwrite AI infrastructure"; according to Huang's post, each of the six institutions will run its own platform and "independently assess each opportunity: the customer, demand, utilization, cash flow and residual value," while NVIDIA "provides the AI factory platform."^[1][3] Huang said on CNBC that all six financiers would make their own lending decisions and that NVIDIA would connect customers with financing partners; he also said the "system architectures are going to be specified in such a way" that "if anything were to happen, somebody else could take it over and operate it."^[5][10]

Bloomberg, citing a person familiar with the plans, gave the most specific description of the intended structure: the deals would use compute as collateral for debt raised through private offerings and through bonds issued by special-purpose entities; those vehicles could issue tens of billions of dollars of debt at a time and then lease the compute to NVIDIA customers; the first deals were expected within months; and the compute would be treated as liquid, so that financing could be reallocated to other buyers if a customer failed.^[6] Executives indicated the effort would focus on debt financing for NVIDIA's largest customers and that many deals already in the works could count toward the commitment.^[6] The Wall Street Journal wrote that "no money has been raised yet under the new partnerships" and that the goal was to sell public and private debt to pension funds, insurers and sovereign wealth funds and use the proceeds to establish the "platforms."^[19] Fortune's Eva Roytburg summarized the chain as an independent vehicle buying GPUs and data-center infrastructure, an AI company leasing or committing to use the compute, and Apollo, KKR and peers structuring the resulting debt for the insurance and retirement capital they manage.^[17]

The one term NVIDIA has quantified is its own participation. Huang wrote that "in some cases, NVIDIA may provide a residual-value support mechanism for up to 25% of an opportunity, assessed carefully on a project-by-project basis," that the support is "limited, residual-value based and designed to complement, not replace, independent underwriting," and that this is "substantially lower than other compute-financing arrangements."^[3] He summarized NVIDIA's role as helping "unlock a very large pool of independent capital while maintaining disciplined risk exposure."^[3][6] Reuters, Forbes and other outlets translated the 25% ceiling into a potential backstop of "up to $125 billion"; Reuters attributed the dollar figure to Huang's X post, but it does not appear in NVIDIA's release or in the blog version of the post, which state the support only as a percentage; 25% of $500 billion is $125 billion.^{[20][25][1][3]} NVIDIA did not share terms, individual commitments or a timeline.^[15] CNBC characterized it as NVIDIA having "the option of backstopping 25% of every loan," which it said should produce lower interest rates for borrowers that had previously depended on their own credit ratings.^[10][28]

Several details remain unspecified. Forbes contributor Robert Szczerba noted on the day of the announcement that there was "no timeline yet, no word on how the money splits among the six firms, and no first project named."^[16] Apollo's Zelter, asked on CNBC about the $35 billion AI-infrastructure financing platform that Apollo, Blackstone and Broadcom had announced in June 2026, called the NVIDIA plan "another example" of compute becoming "a financial asset you could actually fund and finance."^[5][14] Reuters contrasted the NVIDIA structure with earlier AI-infrastructure deals that relied on vendor guarantees, such as Broadcom's residual-value guarantee on roughly $30 billion of senior debt backing Anthropic's chip financing.^[20]

NVIDIA's "investable asset" argument

Huang's post makes the case that NVIDIA compute has the characteristics lenders look for in infrastructure: it produces revenue, serves a broad market, improves over time and can be redeployed. He argued that a DSX AI factory can serve many customers and workloads, that the architecture is used by every major cloud so a factory "can be used by another customer, another cloud or another operator," and that CUDA software "improves the performance, efficiency and total cost of ownership of already-installed infrastructure." As evidence of longevity he cited the A100, introduced in 2020 and "six years later" still in commercial use, with customers committing multi-year capacity and "extending A100's economic life toward a decade."^[3]

He also cited rental-price data, which NVIDIA presented without naming a source:

Scroll sideways for more →

The post then poses and answers four questions. On "Is this circular financing?" it says: "This initiative is designed to address that concern. We are bringing independent, long-term institutional capital into the AI infrastructure market," that "the demand is real," and that "the capital providers independently underwrite each project." On "Why would NVIDIA support financing?" it gives the 25% residual-value mechanism. On "Can the market absorb this capacity?" it says capacity "will be built around real customer economics." On "Where is the return on investment?" it answers, "The return is in the usefulness of AI."^[3] On CNBC, Huang added that "There will always be a customer for that computing platform" because NVIDIA's architecture is "fairly universally adopted."^[5]

Critics read the rental data differently. Szczerba pointed out that a rising one-year rental rate "shows what the chip earns now. It doesn't show what it would sell for," and that by Silicon Data's tracking the median price to rent an H100 from a large cloud fell from about $9.34 an hour in the second half of 2024 to about $6.26 a year later, so "there's no single H100 price." He also noted that Amazon shortened the estimated useful life of some servers from six years to five, effective January 1, 2025, citing "the increased pace of technology development, particularly in the area of artificial intelligence and machine learning," which added about $1.4 billion to its 2025 depreciation.^[16] SiliconANGLE's Dave Vellante drew the same distinction: "Functional life is not the same as economic residual value," and if NVIDIA must provide residual-value support, "the market has not yet accepted the thesis without credit enhancement."^[24]

Market reaction and analyst views

NVIDIA shares fell after the Financial Times report on August 10; Forbes reported them down about 3% around midday and 2.4% (around $218) in afternoon trading, a move it valued at roughly $130 billion of market capitalization.^[7] CNBC reported the shares about 1% higher during trading on August 11 as analyst notes circulated.^[9]

Sell-side reaction split along the lines of the existing circular-financing debate. Bank of America's Vivek Arya wrote that the initiative "appears to be a pivot away from vendor-financing" and that the burden "sits with the consortium" rather than "NVDA's balance sheet," adding that "NVDA guarantees asset quality, not the debt, turning bears' depreciation worry into the enabling feature."^[9][20] Morgan Stanley's Joseph Moore said the platforms should "alleviate circularity concerns," while cautioning that "the view that the AI ecosystem will take on more leverage is in itself an investment debate, even if Nvidia does not provide the leverage." Wells Fargo's trading desk countered that "in the end NVDA is still a part of the financing," calling the agreements a form of "depreciation insurance," and Mizuho traders wrote that the plan "doesn't fundamentally answer the question of how much end-user demand and economic return sits underneath all of this spending."^[9] Hedgeye's Felix Wang told Bloomberg that "in effect, they made Nvidia's product cheaper without really cutting GPU prices," while making "future demand more sensitive to credit conditions."^[6] Axios wrote on the day of the announcement that the move "could reignite fears about the circular nature of AI financing."^[8] A week later, in a note on NVIDIA's Ohio guarantee for OpenAI, Cantor analysts wrote that they viewed the company's financing "less as circular and more facilitating the coming AI buildout."^[28] Forbes contributor Sanjit Singh Dang described the same shift approvingly, as NVIDIA "becoming a capital allocator for the AI era."^[30]

Credit analysts were more guarded. IFR reported that Morgan Stanley credit analysts Lindsay Tyler and Nishant Satyam described NVIDIA's commitments as "balance-sheet-as-a-service," and that NVIDIA's bonds, despite its double-A ratings, traded at spreads closer to a triple-B issuer, which they ascribed to circular-financing concerns; they wrote that "much of the tail risk... remains early stage and difficult to track, with the US$500bn partnership still in MoUs, limited detail on revenue-sharing or credit support, and potentially limited disclosure as financing migrates through SPVs and private structures."^[39] Fortune reported Goldman Sachs' estimate that AI-related financing accounted for nearly a quarter of all gross US investment-grade issuance in 2026, and Alternative Credit Investor cited Morgan Stanley research from May 2026 projecting more than $3.2 trillion of data-center capital spending through 2028, of which about $1.75 trillion would be financed with credit and roughly $700 billion with private credit.^[17][14]

Several commentators noted that the announcement was thin on specifics. CNBC observed that the panel discussion "was thin on specifics as far as the types of borrowers that will emerge, what interest rates will look like, where the facilities will be constructed and when it will all kick off," and recalled that NVIDIA's September 2025 plan to invest up to $100 billion in OpenAI "never materialized" in that form, although NVIDIA later contributed $30 billion to OpenAI's 2026 funding round.^[10] Infrastructure Investor asked whether investors would recognize compute as an infrastructure asset at all.^[29]

Criticism

Scroll sideways for more →

Burry returned to the theme before NVIDIA's earnings, writing on Substack on August 26 that NVIDIA's stock was "wildly undervalued" on its face yet "treading water," which he attributed to the market's doubts about demand built on what he has called "byzantine financing arrangements."^[38] Fortune's Roytburg noted that the capital targeted by the platforms is largely insurance and retirement money, and quoted Bloomberg columnist Matt Levine's summary of the strategy as putting more private investments into ordinary people's retirement accounts, raising "a gazillion dollars" of private-credit and infrastructure funds, and using it to build the data centers AI will rent.^[17]

Regulators had flagged the underlying pattern before the announcement. The Bank of England's July 2026 Financial Stability Report warned that AI companies' revenue forecasts "may reflect 'circular financing arrangements'," in which "technology companies invest in AI companies which in turn purchase those technology companies' products," and that off-balance-sheet structures such as special purpose vehicles and asset-backed vehicles were "broadening the AI debt footprint across leveraged, structured, and private credit" and increasing "the complexity of identifying where risk ultimately sits."^[44] How debt issued through the platforms would be rated has not been disclosed; Morgan Stanley's credit analysts cited the "limited detail on revenue-sharing or credit support" as a reason the risk is hard to track.^[39] Gollakota noted that an $8.5 billion CoreWeave facility secured by GPU infrastructure and a Meta offtake agreement had been rated A3 by Moody's in March 2026, which he cited as evidence that lenders and agencies "care who is ultimately promising to pay."^[43]

Context: NVIDIA's other financing commitments

The platforms were announced in the middle of a run of balance-sheet commitments by NVIDIA that its critics group together under the "circular financing" label, and which NVIDIA disclosed in detail for the first time in its August 26 CFO commentary.^[32][39]

Scroll sideways for more →

IFR summed the disclosed commitments and guarantees to more than $530 billion and reported that the total had more than tripled in three months.^[39] On the earnings call, Bank of America's Arya said his own tally of the CFO commentary came to "about $500 billion or so."^[33]

CNBC had reported in July that NVIDIA was discussing a backstop of up to $250 billion so OpenAI could raise debt for a 10-gigawatt Ohio data center; the August 17 agreement covered less than half that and, according to the NVIDIA release, gives NVIDIA the option to take the remaining 3.75 IT-gigawatts at the site.^[27][26] Huang wrote on X that frontier labs "are growing faster than their balance sheets and long-term credit profiles can support," which CNBC read as the same rationale that underlies the financing platforms.^[28] On August 31, Benzinga, citing the Wall Street Journal, reported that Anthropic had signed a $35 billion cloud-computing agreement with Lambda, an NVIDIA-backed provider, for a Texas data center being developed by Hut 8 where NVIDIA had reportedly secured the capacity; some headlines called this "Nvidia's $35 billion Anthropic pact," although the contract is between Anthropic and Lambda.^[41]

The August 26 earnings call

NVIDIA reported second-quarter fiscal 2027 revenue of $96.2 billion on August 26, 2026, up 106% year on year, and for the first time gave a year-ahead outlook, telling investors that fiscal 2028 revenue growth would be about 70%.^[31][34] The release listed the financing platforms among the quarter's highlights, describing them as intended "to mobilize over $500 billion of third-party capital for the buildout of AI infrastructure over time, subject to definitive agreements."^[31]

On the call, CFO Colette Kress placed the platforms inside NVIDIA's frontier-lab strategy: "to support the Frontier Labs infrastructure build-outs, we recently announced partnerships with six of the world's leading infrastructure capital providers, Apollo, BlackRock, Blackstone, Brookfield, Goldman Sachs and KKR, to establish financing platforms that will raise over $500 billion of third-party capital."^[47] She said that with these partnerships "the AI labs will be able to build and assess AI infrastructure funded by long-term institutional capital at relatively attractive rates."^[33] She also described a separate "revenue-sharing structure" for neoclouds, under which NVIDIA "provides a take-or-pay commitment on a portion of the facility's capacity, a minimum revenue guarantee that gives lenders the confidence to underwrite the project," in exchange for a share of revenue above that floor; "Independent capital still underwrites every deal on its own merits. We're not making loans," she said, adding: "In this model, we get paid twice, once on the hardware sale and again, through the share of rental revenue."^[33][36]

Kress addressed the circularity charge directly: "We recognize the scale of this support, and we know some will call this circular financing. We see it differently. We're going through a major computing platform shift, the creation of one of the most important technologies in human history and these are once-in-a-generation companies."^[47] She said NVIDIA had invested nearly $50 billion in frontier labs, would provide "selective credit enhancement for nearly 2 gigawatts of compute" for a frontier lab other than OpenAI, and expected demand from the labs "for which we expect to leverage our balance sheet to contribute toward roughly a quarter of our business next year," while describing NVIDIA's risk as "limited" because its platform "is fungible and durable and can be redeployed to support other customers."^[33][39] Huang told analysts that "the only regret that I have is that I didn't invest more and sooner."^[33] Days sales outstanding rose to 60 days from 45, which NVIDIA attributed to extended payment terms on large multi-quarter agreements with investment-grade customers.^[32]

The Washington Post's WP Intelligence brief observed that the call "highlighted how the chipmaker is increasingly leveraging its balance sheet to support its customers' ability to buy Nvidia's expensive hardware," while quoting Aptus Capital's David Wagner, an NVIDIA bull, on why the $500 billion platform is "not Nvidia writing itself a check."^[36] CNBC noted that NVIDIA's quarterly filing newly identified indebtedness as a risk factor and that Kress expected the top five hyperscalers' capital spending to rise from $800 billion in 2026 to $1.3 trillion in 2027.^[35]

Timeline

Scroll sideways for more →

Status as of September 4, 2026

The partnerships remain memorandums of understanding. NVIDIA's newsroom carries no release about the platforms after August 10; the only related items are the August 17 SB Energy guarantee and the August 26 earnings release, which again described the platforms as "subject to definitive agreements."^[45][26][31] Apollo's and Blackstone's newsrooms carry only the joint August 10 release, and Brookfield's news page links to Flatt's CNBC appearance rather than a release of its own.^[11][12][46] IFR's Gollakota wrote on September 3 that "nothing that has come out since has filled in the blanks," and Macaskill noted that Goldman had not given a timeline or said whether it had committed a specific amount of capital.^[43][37] Reuters' August 14 report that Goldman was sounding out investors, and Bloomberg's report that the first deals were expected "within months," are the only indications of timing.^[20][6] The NVIDIA Japan X post of September 2 restated the announcement without new information.^[42]

What the platforms will look like in practice, which customers will borrow through them, how the 25% residual-value support will be documented, how the resulting debt will be rated, and whether the vehicles will eventually pool and securitize loans rather than finance projects one at a time are all open questions that NVIDIA and its partners have not answered publicly.^[43][24][37]

References

1. ^NVIDIA Partners With Apollo, BlackRock, Blackstone, Brookfield, Goldman Sachs and KKR to Establish AI Compute Infrastructure Financing Platforms to Mobilize Over $500 Billion of Third-Party Capital - NVIDIA Newsroom, August 10, 2026.
2. ^NVIDIA AI Factory Compute Is Becoming an Investable Asset Class (X article) - X (Jensen Huang), August 10, 2026.
3. ^NVIDIA AI Factory Compute Is Becoming an Investable Asset Class - NVIDIA Blog (Jensen Huang), August 11, 2026.
4. ^Nvidia lines up $500 billion in financing as CEO Jensen Huang tells CNBC his chips are 'investable asset' - CNBC (Hugh Son), August 10, 2026.
5. ^CNBC Exclusive: Transcript: CNBC's Becky Quick Speaks with Nvidia's Jensen Huang & Wall Street Leaders on $500B AI Infrastructure Push on "Closing Bell: Overtime" Today - CNBC, August 10, 2026.
6. ^Nvidia taps Wall Street for $500 billion funding commitment - Bloomberg, syndicated by Fortune (Silla Brush, Ian King, Todd Gillespie), August 11, 2026.
7. ^Nvidia Stock Loses $130 Billion In Market Value As Firm Reportedly Enters $500 Billion AI Financing Deal - Forbes, August 10, 2026.
8. ^Nvidia and Wall Street partner on $500B AI financing - Axios (Nathan Bomey), August 10, 2026.
9. ^Nvidia is trying to quiet 'circular financing' accusations. Wall Street is unsure it will - CNBC (Tobias Burns), August 11, 2026.
10. ^Wall Street just endorsed Jensen Huang's 'big concept' for AI. What now? - CNBC, August 11, 2026.
11. ^NVIDIA Partners with Apollo, BlackRock, Blackstone, Brookfield, Goldman Sachs and KKR to Establish AI Compute Infrastructure Financing Platforms to Mobilize Over $500 Billion of Third-Party Capital - Apollo Global Management press release, August 10, 2026.
12. ^NVIDIA Partners with Apollo, BlackRock, Blackstone, Brookfield, Goldman Sachs and KKR to Establish AI Compute Infrastructure Financing Platforms to Mobilize Over $500 Billion of Third-Party Capital - Blackstone press release, August 10, 2026.
13. ^Nvidia's Risky Business - Stratechery (Ben Thompson), August 11, 2026.
14. ^Nvidia taps US alts giants for $500bn AI financing - Alternative Credit Investor, August 11, 2026.
15. ^Nvidia guarantees its own chips' value to unlock $500 billion in AI infrastructure financing - The Decoder, August 11, 2026.
16. ^Nvidia's $500 Billion Bet To Make AI Compute Wall Street's Next Asset Class - Forbes (Robert J. Szczerba), August 10, 2026.
17. ^Nvidia found a new way to keep the AI boom funded: your retirement money - Fortune (Eva Roytburg), August 12, 2026.
18. ^Michael Burry Calls Nvidia's $500 Billion AI Financing Push a 'Wall Street Stunt': 'Meet the New Boss...' - Benzinga via Yahoo Finance (Rishabh Mishra), August 13, 2026.
19. ^Why Wall Street and Nvidia are building an exotic money pipeline for the AI boom - The Wall Street Journal, republished in Mint, August 13, 2026.
20. ^Goldman in talks with investors on Nvidia financing deal after landing prized role, sources say - Reuters via Yahoo Finance, August 14, 2026.
21. ^Goldman Sachs is courting investors for its lead role in Nvidia's $500B AI financing push - Quartz, August 14, 2026.
22. ^Post on the NVDA $500B consortium plan - X (Jeffrey Gundlach), August 15, 2026 (00:11 UTC).
23. ^'Bond King' Jeff Gundlach Compares Nvidia's $500 Billion AI Financing Push to Bonds Backed by Bananas: 'Will Not Age Well' - Benzinga via Yahoo Finance, August 2026.
24. ^Did Nvidia's Jensen Huang just make the AI buildout too big to fail? - SiliconANGLE (Dave Vellante), August 15, 2026.
25. ^Nvidia AI Financing Is The $500 Billion Risk Investors Aren't Watching - Forbes (Jim Osman), August 16, 2026.
26. ^NVIDIA Guarantees SB Energy's PORTS-Pike Technology Campus in Ohio to Exclusively Host NVIDIA AI Compute - NVIDIA Newsroom, August 17, 2026.
27. ^Nvidia backing $105 billion in financing for OpenAI data center in Ohio - CNBC, August 17, 2026.
28. ^Nvidia's AI moat is shifting from chips to capital - CNBC, August 18, 2026.
29. ^Nvidia's new 'investable asset class' must retain the 'infra' in AI infra - Infrastructure Investor (Zak Bentley), August 19, 2026.
30. ^Nvidia As The New Banker For AI Gold Rush - Forbes (Sanjit Singh Dang), August 22, 2026.
31. ^NVIDIA Announces Financial Results for Second Quarter Fiscal 2027 - NVIDIA Newsroom, August 26, 2026.
32. ^CFO Commentary on Second Quarter Fiscal 2027 Results - NVIDIA Investor Relations (Colette Kress), August 26, 2026.
33. ^Nvidia (NVDA) Q2 2027 Earnings Call Transcript - The Motley Fool via Yahoo Finance, call of August 26, 2026.
34. ^Nvidia gave its first-ever year-ahead forecast - a 70% growth bombshell meant to silence AI bubble critics and 'circular financing' doomsayers - Fortune (Amanda Gerut), August 26, 2026.
35. ^Nvidia earnings takeaways: Huang forecasts 70% fiscal 2028 revenue growth, far above estimates - CNBC (Kif Leswing, Ari Levy, Tobias Burns), August 26, 2026.
36. ^AI & Tech Brief: Nvidia's financing spree - The Washington Post, WP Intelligence (Benjamin Guggenheim), August 27, 2026.
37. ^Macaskill on Markets: Is Nvidia a problem for JP Morgan, or Goldman Sachs? - International Financing Review (Jon Macaskill), August 27, 2026.
38. ^Michael Burry says AI is a 'byzantine' money loop - Nvidia's monster results could be ammo for his argument - Moneywise via AOL, August 27, 2026.
39. ^Nvidia defends circular finance deals as commitments surpass US$530bn - International Financing Review, August 28, 2026.
40. ^NVIDIA and MediaTek Deepen Long-Standing Partnership to Build AI Edge to Cloud Computing Platforms - NVIDIA Newsroom, August 31, 2026.
41. ^Anthropic Signs $35 Billion Cloud Agreement With Nvidia-Backed Lambda in Deal Where Chip Giant Holds the Lease: Report - Benzinga, August 31, 2026.
42. ^NVIDIA Japan post restating the financing-platform announcement - X (@NVIDIAJapan), September 2, 2026.
43. ^Nvidia's US$500bn plan to fill the AI financing gap is wishful thinking - International Financing Review (Prasad Gollakota), September 3, 2026.
44. ^Financial Stability Report - July 2026 - Bank of England, July 2026.
45. ^NVIDIA Newsroom: latest news - NVIDIA Newsroom, accessed September 4, 2026.
46. ^News - Brookfield, accessed September 4, 2026.
47. ^NVIDIA Corp. (NVDA) Q2 2027 Earnings Call, Corrected Transcript - FactSet CallStreet, hosted by NVIDIA Investor Relations, August 26, 2026.