NVIDIA CEO Jensen Huang has personally stepped forward to defend the company's plan to mobilize over $500 billion in AI infrastructure financing alongside six major financial institutions. He clarified that NVIDIA's AI factory computing power is becoming an investable asset class, with demand stemming from real-world business scenarios. Independent institutional investors will conduct due diligence on each project, countering external criticism of "circular financing."
On August 10, Huang announced on X that NVIDIA had partnered with Apollo, BlackRock, Blackstone, Brookfield, Goldman Sachs, and KKR to establish an independent financing platform. This platform aims to mobilize over $500 billion in third-party capital over time to support AI infrastructure development. He emphasized that the $500 billion figure represents the total third-party capital the platform is designed to mobilize, "neither NVIDIA's revenue, nor a single fund or commitment to a single client."
Following the announcement, NVIDIA's stock price briefly fell 3.2%. The company's five-year credit default swap (CDS) price, a measure of credit risk, rose to 77.215 basis points on Monday, up about 5.3 basis points from the previous trading day, marking the largest single-day increase in two weeks. Analysts suggest this reflects growing market concern over NVIDIA's potential credit risk under such a massive financing model.
Earlier reports from financial media indicated that NVIDIA is seeking to join forces with Wall Street giants like Apollo, BlackRock, Brookfield, Goldman Sachs, and KKR to raise up to $500 billion for AI infrastructure projects, including AI chip procurement, power production, and data center construction.
In response to these reports, famed short-seller Jim Chanos, also known as "Dr. Doom," posted a sarcastic comment on social media. He compared NVIDIA's joint financing with financial giants to the financial engineering of the 2008 financial crisis, hinting that if the AI bubble bursts, those involved could face the same scrutiny as Wall Street executives did before Congress.
AI Factories: From 'Buying Chips' to 'Financable Infrastructure'
Huang laid out the asset logic behind NVIDIA's AI factories, aiming to redefine market perception of this business model. He stated that the AI industry has moved from an era of "companies purchasing chips project-by-project and building their own data centers" to a new phase where "AI factories can be financed as productive infrastructure." These factories feature replicable platforms, long-term institutional capital support, and a diverse customer base generating revenue from computing power.
Huang emphasized that NVIDIA's computing power is not just chips but a complete AI factory platform, encompassing accelerated computing, networking, system software, AI frameworks, and a global developer ecosystem. He noted that a single NVIDIA AI factory can serve multiple clients and workloads simultaneously, offering flexibility and substitutability. When a client's needs change, the factory can be repurposed for another client, cloud service provider, or operator. "This broad ecosystem gives NVIDIA's computing power a deep potential user market, helping to protect residual value."
Huang characterized this partnership as "the beginning of an open capital market for AI infrastructure." He said these financial institutions, as global leaders in infrastructure investment, bring deep expertise in underwriting long-term productive assets. Together, they will create a replicable financing platform to support the factories needed for the AI ecosystem.
CUDA Software Continues to Appreciate, A100 Still in Commercial Use After Six Years
Huang stressed the long-term economic value of NVIDIA's computing assets, backing it with specific data. He explained that CUDA software continuously improves the performance, efficiency, and total cost of ownership of installed infrastructure, enabling AI factories to produce more intelligence at lower costs throughout their lifecycle, extending their economic useful life.
Using the A100 as an example, Huang noted that NVIDIA launched the Ampere-architecture-based A100 in 2020. Six years later, it remains actively used in commercial applications for AI training, fine-tuning, inference, and high-performance computing, with clients signing multi-year capacity contracts. "The A100's economic life is extending toward a decade."
Regarding GPU leasing pricing, Huang cited market data showing that the one-year lease price for the H100 has risen from about $1.70 per GPU hour in October 2025 to approximately $2.35 in March 2026. The median on-demand pricing across service providers increased from about $2.00 in October 2025 to $2.70 in June 2026. The Blackwell series commands a higher premium, with B200 cloud pricing ranging from $5.30 to $7.05 per GPU hour. He argued these figures confirm the enduring economic value of NVIDIA's computing power.
Directly Addressing 'Circular Financing' Concerns
In response to the most prominent market concern about "circular financing," Huang dedicated a Q&A section in his post. He said the financing structure was designed specifically to address this worry. Demand comes from frontier AI labs, AI-native startups, enterprise clients, cloud service providers, and countries building AI services. "The demand is real." Each capital source will independently conduct due diligence on each project, evaluating factors like client qualifications, demand conditions, utilization rates, cash flow, and residual value. "NVIDIA provides the platform; investors make independent financing decisions."
Regarding NVIDIA's own risk exposure, Huang disclosed that in some cases, NVIDIA may provide a residual value support mechanism of up to 25% for a single project, assessed on a project-by-project basis. He stressed that this percentage is "far lower than other computing power financing arrangements" and is a residual value support, designed to supplement, not replace, independent due diligence.
Huang concluded his post by placing AI factory construction within a broader historical framework. He noted that every industrial revolution was built on infrastructure—electricity, transportation, communication, and computing—and each relied on external financing. "AI factories are the infrastructure of the intelligence age."
He summarized the business logic of AI as a positive cycle: businesses use AI to write software, develop drugs, design products, serve customers, automate operations, and build new services. More computing power leads to better AI, better AI leads to more usage, more usage leads to more revenue, and more revenue drives more investment in computing power. "This is the virtuous cycle of the AI industrial revolution."
Huang stated that through this partnership, NVIDIA and the world's leading financial institutions will jointly finance the infrastructure for this industrial revolution, making AI factories more accessible to businesses, industries, and nations.
This is not NVIDIA's first deep involvement in AI industry chain financing. Earlier reports suggested NVIDIA is in talks to provide up to $25 billion in financing guarantees for OpenAI and is discussing a $350 billion financing plan for OpenAI's chip procurement. Last month, NVIDIA also announced an expanded partnership with South Korea's SK Group, with a combined business scale exceeding $500 billion. These moves indicate NVIDIA is transforming from a chip supplier into a core capital mobilization engine for the AI infrastructure ecosystem.
'Dr. Doom' Compares to 2008: Next Congressional Hearing?
Around the same time as Huang's post, well-known Wall Street short-seller James Chanos, also called "Dr. Doom," posted a pointed comment on social media. He wrote: "The next time these people sit together to explain AI financing, it might not be until a 2031 congressional hearing..."
Chanos's implication referenced the 2008 financial crisis, when executives from major Wall Street institutions were forced to testify before Congress over systemic risks caused by subprime mortgages, CDS, and CDOs. He compared NVIDIA's joint financing arrangement with Apollo, BlackRock, Blackstone, Brookfield, Goldman Sachs, and KKR to the financial engineering of that era, suggesting that if the AI infrastructure investment bubble bursts years later, Huang and the executives of these financial institutions could face congressional inquiries. Notably, when asked in comments whether those involved would face criminal charges, Chanos clearly stated: "No one said anything about going to jail."
Comments