- Nvidia is in advanced talks to provide roughly $250 billion in financial guarantees to backstop OpenAI’s lease obligations on a 10-gigawatt data-center complex in southern Ohio — a commitment of unprecedented scale in which the chipmaker would essentially guarantee OpenAI’s ability to fulfill a long-term lease, providing the financial credibility necessary to attract the capital required for a project of this magnitude; the 10-gigawatt figure represents an extraordinary level of computing power, dwarfing existing data-center campuses that typically measure in the hundreds of megawatts, and the Ohio project alone would consume more electricity than many mid-sized cities; the backstop arrangement reflects the degree to which Nvidia’s financial interests are now inseparable from the buildout of AI infrastructure at the frontier.
- SoftBank’s energy subsidiary is the developer behind the Ohio complex, continuing the Japanese conglomerate’s aggressive expansion into AI infrastructure after its $100 billion Stargate commitment earlier this year; the total cost of the project — including the Nvidia chips that will be installed inside the data centers — could exceed $500 billion, making it the single largest data-center project announced to date by a wide margin; the scale reflects the AI industry’s accelerating conviction that frontier model training and inference will require computing resources that dwarf anything previously built, and that securing those resources now, before demand fully materializes, is the critical competitive advantage in the race to AGI.
- For Nvidia, providing a $250 billion financial guarantee represents a remarkable evolution from chip supplier to strategic financier of the AI ecosystem it has built — by backstopping OpenAI’s lease, Nvidia effectively guarantees both the revenue stream from chip sales embedded in the project and the continued scaling of the infrastructure that drives demand for future GPU generations; this financial engineering mirrors what hyperscalers like Microsoft and Google have done in securing long-term cloud capacity, but with Nvidia in the unusual position of financing the customer rather than the other way around; the arrangement also gives Nvidia a form of strategic control over one of the most consequential AI deployments in history.
- The $500 billion-plus project scope puts it in the same category as national infrastructure investments rather than corporate capital expenditures, and the involvement of U.S. government-controlled power resources underscores the degree to which AI infrastructure has become a matter of national strategic interest; the Ohio site selection reflects that state’s emerging role as a data-center corridor due to relatively low electricity costs, available land, and grid access; if completed, the Ohio complex would reshape the global AI competitive landscape by giving OpenAI a compute advantage that would be extraordinarily difficult for any private competitor to match or exceed.
What Happened?
Nvidia is in talks to provide approximately $250 billion in financial guarantees to underwrite OpenAI’s lease of a 10-gigawatt AI data-center complex in southern Ohio being developed by SoftBank’s energy subsidiary, according to people familiar with the matter. The total project cost, including the Nvidia chips to be installed inside, could exceed $500 billion — making it the largest data-center project ever announced. The deal would involve power resources controlled by the U.S. government, reflecting the national-security dimensions of frontier AI infrastructure.
Why It Matters?
The scale of this transaction signals a fundamental shift in how AI infrastructure is being financed and structured. Nvidia moving from chip supplier to financial backstop for a $250 billion obligation is without precedent in the tech industry, and suggests that the economics of frontier AI now require vertical integration not just in hardware and software but in capital markets as well. For OpenAI, securing a 10-gigawatt campus would provide a training and inference advantage that could cement its position at the frontier for years. For the broader AI industry, the project resets expectations about what “large” means in the context of AI infrastructure investment.
What’s Next?
The deal remains in talks and has not been finalized; key variables include the structure of Nvidia’s guarantee obligations, the financing terms for the SoftBank-developed real estate, and the power agreements with government-controlled utilities. Given the $500 billion-plus total price tag, this project will require participation from a syndicate of lenders, sovereign wealth funds, and potentially government financing vehicles — any of which could delay or reshape the final structure. Watch for announcements around power supply agreements and construction timelines as the clearest signal that the deal is moving from negotiation toward execution.
Source: The Wall Street Journal
