AIToday
AI Business & IndustryAI Stocks & MarketsYahoo Finance AIPublished: Aug 17, 2026, 19:00 JST2 min read

Nvidia, OpenAI trim Ohio data center deal to under $120B

Nvidia, OpenAI trim Ohio data center deal to under $120B

Key takeaway

  • Nvidia and OpenAI are finalizing a deal to build a data center in Ohio, with Nvidia reducing its financial guarantee to less than $120 billion from $250 billion originally.

  • Nvidia will finance the project's first 5-gigawatt phase and decide later whether to fund additional capacity, while OpenAI is negotiating a binding lease for 10 GW.

  • The restructured arrangement lets both companies proceed with major AI infrastructure while lowering Nvidia's upfront risk.

3 Key Points

  1. What happened

    Nvidia and OpenAI are finalizing a deal to finance a data center in Ohio, with Nvidia's financial guarantee reduced to less than $120 billion from an initially planned $250 billion. Nvidia would back the project's initial 5-gigawatt phase and decide later on financing the rest, while OpenAI is in talks to sign a binding lease for 10 GW in the coming days.

  2. Why it matters

    The restructured deal signals both companies are moving forward with major AI infrastructure investments while managing financial exposure. For Nvidia, a lower upfront guarantee reduces risk; for OpenAI, securing a 10 GW lease locks in computing capacity for training and running its AI models—essential as demand for data center resources intensifies.

  3. What to watch

    Whether OpenAI signs the binding lease in the coming days, as the Wall Street Journal reported; the success of the initial 5-gigawatt phase will determine Nvidia's willingness to finance the remaining capacity.

Ask the AI about this article →

Context & Analysis

The reworking of the Nvidia-OpenAI Ohio data center deal reflects a pragmatic shift in how the two companies are structuring large infrastructure commitments. Originally planned with Nvidia guaranteeing $250 billion, the reduction to less than $120 billion suggests both parties sought to lower risk exposure while maintaining momentum on a strategically critical project. By phasing Nvidia's commitment—supporting the initial 5 gigawatts upfront with a future decision point on additional capacity—the deal creates a validation period: the success or performance of the first phase will inform whether Nvidia commits to larger backing.

For OpenAI, the parallel negotiation of a 10 GW binding lease is the company's way of securing long-term computing capacity independent of Nvidia's financing scope. Data center power has become the primary constraint on AI model training and deployment, so locking in this lease ensures OpenAI can run its operations and serve its customers without competing for limited electricity and infrastructure. The deal reflects the reality that AI companies need guaranteed access to enormous computational resources, and such access now requires direct involvement from hardware makers like Nvidia.

FAQ

How much will Nvidia's guarantee be under the new deal?
Less than $120 billion, down from the initially planned $250 billion.
What capacity is Nvidia committing to in the first phase?
Nvidia would support the project's initial phase, totaling around 5 gigawatts, with a decision on financing the rest to come later.
How much power has OpenAI agreed to lease?
OpenAI is in discussions to sign a binding lease for 10 GW of the project in the coming days.
Yahoo Finance AIRead Original Article

Get the latest AI Business & Industry news every morning

For example, today's edition would include:

  • Visko raises $10M, launches live AI video model OrbisSiliconANGLE AI · 1h ago
  • NVIDIA Leads 5 AI Stocks to Watch in SeptemberYahoo Finance AI · 1h ago
  • UAE AI firm pitches 95% video compression for Saudi securitySemafor Tech · 1h ago

AI-summarized, only the topics you pick — one digest a day via Email, Slack, or Discord.

Free · takes 30 seconds · unsubscribe anytimeWhat is AIToday? →

Ask AI

Ask AI anything about this article. Q&As are published on this page for other readers too.

Related Articles

Next articleAI-generated wildlife images pose risks to animals and conservation, experts warn