AI

Nvidia in Talks to Provide $250 Billion Guarantee for OpenAI Data Center in Ohio

Nvidia is negotiating a $250 billion payment guarantee to help OpenAI lease computing capacity from a 10 GW AI data center project in Ohio developed by SoftBank.

By Tim Editorial

Nvidia in Talks to Provide $250 Billion Guarantee for OpenAI Data Center in Ohio
img.datacenterfrontier.com

Nvidia is negotiating to provide a payment guarantee worth approximately $250 billion to back OpenAI's lease of computing capacity from a massive artificial intelligence data center project in Ohio, according to a report published over the weekend. If realized, the guarantee would rank among the largest financial commitments ever made by a technology company to an AI infrastructure project. The project, developed by SoftBank, is planned to become one of the world's largest AI computing facilities, with a power capacity of 10 gigawatts. That is equivalent to the electricity consumption of roughly 8 million to 10 million U.S. households, or the output of several nuclear power plants. By comparison, current large scale AI data centers typically operate at tens to hundreds of megawatts.

The 10 GW scale underscores an ambition to build unprecedented AI infrastructure. The U.S. government is involved in controlling the electricity supply for the project, indicating it may fall under national energy security oversight. The enormous power demand is required to run thousands of Nvidia graphics processing units that will train and operate next generation AI models. For Nvidia, the move expands its role from chip supplier to financial infrastructure partner. By providing a backstop, Nvidia would effectively assume the risk that OpenAI might fail to pay its computing lease in the future. This is an unusual strategy for a chip company, which typically does not guarantee property or infrastructure lease payments.

OpenAI, as the primary user of AI computing, requires vast and stable capacity to develop increasingly sophisticated models. Successors to GPT 4 are expected to require far more computing power than their predecessors. With Nvidia's guarantee, OpenAI can secure access to necessary infrastructure without committing large upfront capital. SoftBank, through its Vision Fund, has long been a major backer of AI and technology companies. Its involvement as developer of the Ohio project signals a long term commitment to physical AI infrastructure, not just equity investments. Ohio has become an attractive location for data centers due to land availability, tax incentives, and access to a stable power grid. However, a project of this scale faces significant regulatory and technical challenges.

Building a 10 GW data center requires massive investment in electrical infrastructure, including substations, transmission lines, and possibly new power plants. Environmental and construction permitting will be lengthy and complex. Neither Nvidia, OpenAI, nor SoftBank have officially commented on the report. Sources cited by the Wall Street Journal emphasized that negotiations are preliminary and could change or collapse. There is no guarantee that a final deal will be reached at the reported value or structure. If the deal materializes, its impact would ripple across the AI ecosystem. First, it would strengthen Nvidia's position as the dominant AI chip supplier while locking in long term demand for its GPUs. Second, it would provide OpenAI with nearly unlimited computing resources for research and development.

Third, it would set a precedent for other technology companies to invest directly in physical AI infrastructure with financial backing from partners. The project could also spark new competition among U.S. states to attract AI data center investments. Ohio competes with states like Virginia, Texas, and Arizona, which also have competitive power infrastructure and incentives. The federal government, under energy security oversight, will likely play a key role in accelerating or hindering such projects. From a market perspective, the news comes amid growing concerns about AI data center energy consumption. Some analysts estimate that AI data centers could consume up to 10% of global electricity by 2030. The 10 GW Ohio project would be one of the largest contributors to that trend.

Industry and financial observers will closely watch for further developments. If negotiations succeed, an official announcement could come within months. However, given the complexity and scale, full operation of the data center may take years. The report first emerged late last week, detailing that Nvidia is in talks to provide a backstop or payment guarantee of about $250 billion. The guarantee is intended to cover the cost of computing capacity that OpenAI would lease from the data center. SoftBank, a major global technology investor, is acting as the primary developer of the Ohio project. For context, the 10 GW capacity is equivalent to the electricity consumption of approximately 8 to 10 million U.S. households, or comparable to the output of several nuclear power plants.

Current large scale AI data centers typically operate at capacities of tens to hundreds of megawatts. The 10 GW project demonstrates an ambition to build AI infrastructure on an unprecedented scale. The involvement of the U.S. government in controlling the electricity supply is a key factor. It suggests the project may be under national energy security supervision, given the enormous electricity demand required to run thousands of Nvidia GPUs that will be used to train and run next generation AI models. For Nvidia, this step represents an expansion from being merely a chip supplier to a financial infrastructure partner. By providing a guarantee, Nvidia effectively assumes the risk that OpenAI may not be able to pay the computing lease in the future.

This is an unusual strategy because chip companies typically do not directly guarantee property or infrastructure lease payments. OpenAI, as the primary user of AI computing, needs vast and stable computing capacity to develop increasingly sophisticated AI models. Models like GPT 5 or its successors are expected to require significantly more computing power than their predecessors. With Nvidia's guarantee, OpenAI can secure access to the necessary infrastructure without having to spend large amounts of capital upfront. SoftBank, through its Vision Fund, has long been a major supporter of AI and technology companies. Their involvement as the developer of the Ohio project demonstrates a long term commitment to physical AI infrastructure, not just equity investments.

Ohio itself has become an attractive location for data centers due to land availability, tax incentives, and access to a stable power grid. However, a project of this scale will certainly face regulatory and technical challenges. Building a 10 GW data center requires massive investment in electrical infrastructure, including the construction of substations, transmission lines, and possibly new power plants. Environmental and construction permits will also be a long and complex process. Neither Nvidia, OpenAI, nor SoftBank have provided official comments on the report. Sources cited by the Wall Street Journal emphasized that negotiations are still preliminary and could change or fail. There is no guarantee that a final deal will be reached with the same value or structure as reported.

If this deal materializes, its impact will be felt across the AI ecosystem. First, it will strengthen Nvidia's position as the dominant supplier of AI chips, while locking in long term demand for their GPUs. Second, it will provide OpenAI with nearly unlimited computing resources for research and development. Third, it will set a precedent for other technology companies to invest directly in physical AI infrastructure with financial support from partners. This project could also trigger new competition among U.S. states to attract AI data center investments. Ohio competes with other states like Virginia, Texas, and Arizona, which also have competitive power infrastructure and incentives. The federal government, under energy security oversight, will likely play an important role in accelerating or hindering such projects.

From a market perspective, this news comes amid growing concerns about the energy consumption of AI data centers. Some analysts estimate that AI data centers could consume up to 10% of global electricity by 2030. The 10 GW project in Ohio would be one of the largest contributors to this trend. Further developments will be eagerly awaited by the technology and financial industries. If negotiations are successful, an official announcement could occur within the next few months. However, given the complexity and scale, the process could take years before the data center is fully operational.

Sources and references