Sam Altman had a deadline — and a $100 billion deal to close.
In a whirlwind of late-night calls, hurried meetings, and presidential briefings, OpenAI and Nvidia struck one of the largest partnerships in tech history, hours before Altman boarded a flight to Texas to unveil new AI infrastructure plans.
Key Takeaways
- Nvidia commits $100B to OpenAI in staged $10B tranches.
- Altman and Huang closed terms after Trump’s U.K. state visit.
- Microsoft informed only a day before the deal was signed.
- Stargate project now covers all OpenAI infrastructure builds.
- OpenAI may take on debt, explore becoming a cloud provider.
OpenAI and Nvidia sealed a $100 billion deal after urgent CEO talks, finalizing terms during Trump’s U.K. visit. The pact deepens OpenAI’s reliance on Nvidia’s chips while fueling massive new AI data centers under the “Stargate” project. Microsoft was informed only a day before the agreement closed.
A $100 Billion Pact Sealed in Urgency
Hours before Sam Altman flew to Abilene, Texas, to showcase OpenAI’s next infrastructure push, he and Nvidia CEO Jensen Huang locked in one of the largest tech partnerships ever. Through late-night calls and a flurry of meetings across London, Washington, and San Francisco, the two leaders closed a $100 billion deal that will define AI’s next chapter.
The announcement came Monday at Nvidia’s Silicon Valley headquarters, where Altman told CNBC: “We have to figure out how to do this unprecedented infrastructure challenge.”
Behind-the-Scenes Negotiations
The talks had been weeks in the making but gained momentum during President Donald Trump’s state visit to the U.K. Both Altman and Huang attended, briefing the president in advance of the final signing. No bankers were involved, underscoring the direct nature of the negotiations.
Why This Deal Matters
The pact binds together the world’s most valuable public company and the most prominent AI startup. Nvidia, now worth about $4.5 trillion, surged by $170 billion on news of the deal. OpenAI, already valued at $500 billion, gains guaranteed access to the chips powering frontier AI models and next-gen data centers.
The financing is structured in $10 billion tranches, with nine more to follow as infrastructure builds come online. Crucially, this limits dilution and keeps OpenAI in control.
Microsoft, Oracle, and Other Partners
Notably, Microsoft — OpenAI’s largest shareholder — learned of the deal just a day before it was signed. This marks a shift from earlier years when Microsoft was OpenAI’s exclusive cloud provider.
Meanwhile, Oracle has separately secured a $300 billion, five-year commitment starting in 2027, under the umbrella of “Stargate,” a megaproject backed by Oracle, SoftBank, and the Trump administration. All of OpenAI’s new buildouts, including the Nvidia partnership, now fall under Stargate.
Inside the Numbers
- $100B — Nvidia’s total planned commitment.
- $10B — Size of each tranche, pegged to valuation at closing.
- 10 gigawatts — Capacity of planned sites, first online late 2026.
- 700–800 sites — Number of locations reviewed since January.
Industry Response
For Nvidia, this deal caps a year of aggressive expansion. Alongside OpenAI, it invested $5B in Intel for chip co-development and $700M in U.K. startup Nscale. Analysts describe Nvidia’s moves as consolidating power across the AI supply chain (Financial Times, 2025).
OpenAI, on the other hand, is signaling more independence. Executives hinted at operating its own cloud services — a move that would put it in direct competition with Amazon Web Services, Microsoft Azure, and Google Cloud.
The Bigger Picture
The partnership is not exclusive. OpenAI executives emphasized they’ll continue working with other chipmakers and cloud providers to avoid being locked into a single vendor. But Nvidia remains the “preferred” partner, giving it unmatched leverage over the most ambitious AI projects on the horizon.
Future Outlook
Executives suggest that within one to two years, OpenAI may launch its own commercial cloud offering. For now, demand for training GPT-class models consumes nearly all available compute. Financing will include debt in addition to equity, highlighting the sheer cost of AI at scale.
The broader AI race is reshaping geopolitics, finance, and infrastructure. The Nvidia-OpenAI pact crystallizes a new power center — one where compute capacity, not just algorithms, dictates competitive advantage.
Conclusion
The Altman–Huang deal is more than just a funding round. It cements Nvidia and OpenAI at the core of global AI infrastructure, setting the stage for trillion-dollar battles over compute, cloud, and control of the industry’s future.