OpenAI weighs leasing 10-gigawatt Ohio data center with Nvidia backing

OpenAI is negotiating to lease a 10-gigawatt data center campus on federal land in Ohio, with Nvidia considering financial backing for the deal.

OpenAI is in talks to lease a proposed 10-gigawatt data center campus on federal land in Ohio, a deal that could include financial backing from Nvidia, The Information reported Tuesday. The 10-GW capacity would make it one of the largest known AI infrastructure projects in the US.

The Information reported the negotiations Tuesday, citing two people with direct knowledge of the discussions. Reuters could not immediately verify the report.

A 10-gigawatt facility would consume enough electricity to power roughly 8 million US homes, based on average residential consumption rates. Typical hyperscale data centers operate at 100 to 300 megawatts, making this project 30 to 100 times larger than a standard facility. The campus would be built on federal land, potentially streamlining permitting and grid interconnection through the PJM Interconnection market.

The scale of the proposed campus reflects the accelerating electricity demands of training and running frontier AI models. Nvidia's potential financial backing would represent an unusual direct investment in data center real estate for the chipmaker, which has primarily focused on selling graphics processing units and networking hardware. For investors tracking the AI infrastructure buildout, the project adds to a growing pipeline of multi-gigawatt campuses that cloud providers and AI companies are racing to bring online.

The Ohio project comes as AI companies and hyperscalers compete for access to power-constrained data center capacity. Microsoft, Amazon, and Google have each announced plans for multi-gigawatt data center campuses across the US, with power availability emerging as the primary bottleneck for AI expansion. The Philadelphia Semiconductor Index has gained 68% this quarter, reflecting investor enthusiasm for the AI hardware spending cycle.

Nvidia's role as a potential financial backer, rather than just a hardware supplier, could reshape the economics of AI infrastructure. The company's graphics processing units power the majority of AI training and inference workloads, and co-investing in data centers would give Nvidia a stake in the long-term capacity buildout. The chipmaker's data center revenue has grown more than 200% year over year in recent quarters, driven by demand from OpenAI, Microsoft, and other AI developers.

For OpenAI, securing dedicated compute capacity at this scale would reduce its reliance on third-party cloud providers and give the company more control over its infrastructure costs. The ChatGPT maker has been expanding its direct data center footprint as it develops more capable AI models that require exponentially more computing power.

The broader AI chip market is also seeing increased activity. Cerebras Systems, which makes wafer-scale chips for AI inference, surged more than 17% Monday after nine Wall Street firms initiated coverage with bullish ratings. Morgan Stanley set a $250 price target on the stock, while Citigroup assigned a $340 target, citing growing demand for low-latency inference. Cerebras reported a backlog of $24.6 billion at the end of 2025, with a significant portion tied to a cloud-services agreement involving OpenAI.

"As AI workloads become increasingly reasoning-intensive, demand for fast, low-latency inference is growing rapidly," Morgan Stanley analysts led by Joseph Moore wrote in a research note.

This article is for informational purposes only and does not constitute investment advice.