NVIDIA’s $100B Bet on OpenAI Now Sees a Chip Leasing Twist

At this point, there’s a lot going on between OpenAI and NVIDIA. Last week, NVIDIA confirmed it will invest a whopping $100 billion into OpenAI. That amount is undoubtedly one of the biggest investments ever in artificial intelligence.

Now, according to The Information, the two companies are working on a different kind of arrangement. Per the report, OpenAI plans to lease NVIDIA’s chips and spread the cost over time.

AI firms usually acquire chips up front to build massive data centers. But, with this move, OpenAI plans to lower the financial pressure as it trains and deploys larger models.

NVIDIA CEO Jensen Huang has said that building an AI data center with one gigawatt of capacity can cost around $50 billion, with about $35 billion of that going toward NVIDIA GPUs. Under the lease structure, OpenAI avoids that steep upfront bill while still scaling up.

The deal also includes $10 billion in near-term funding, with the rest of the money set to follow in phases. To catch you up, OpenAI recently reported $12 billion as its annual revenue and expects as much as $20 billion next year. But it is still operating at a loss.

Besides NVIDIA, OpenAI’s partnership with Microsoft is also taking some turns these days. Microsoft has started exploring other AI companies. Yesterday, the company officially announced adding Anthropic’s Claude AI to Copilot.

That’s not all; a report from earlier this month also hinted that Microsoft is preferring Claude Sonnet 4 AI models over OpenAI’s GPT-5 in VS Code.

Via: Reuters