The talks centre on a planned 10-gigawatt complex being developed by SB Energy, the renewable energy and infrastructure arm controlled by SoftBank Group. The project could require more than $500 billion of total investment, making it one of the largest data centre proposals yet disclosed.
No final agreement has been reached, and the structure, scale and timing of Nvidia’s backing could change. The proposed guarantee would not mean Nvidia directly paying $250 billion towards construction. Instead, it could provide lenders with protection against OpenAI failing to meet obligations under a long-term lease, helping the project raise debt on more favourable terms.
OpenAI does not carry an investment-grade credit rating, creating a challenge for developers seeking to borrow hundreds of billions of dollars against its future rental commitments. Nvidia’s balance sheet and central position in the AI industry could strengthen lender confidence and lower the risk premium attached to the transaction.
The first stage of the Ohio campus is expected to provide about 800 megawatts of capacity by 2028. The full development would consume as much power as several large cities and require extensive generation, transmission, cooling and networking infrastructure alongside the construction of computing halls.
Nvidia is also discussing support for the purchase of processors and related equipment that could cost about $350 billion. That financing would be separate from the proposed lease guarantee, although both elements would reinforce demand for Nvidia’s graphics processing units and networking systems.
The discussions illustrate how the financing of AI infrastructure is becoming increasingly intertwined with the companies supplying its technology. Nvidia has already invested in cloud-computing operators, data centre ventures and AI developers that are major purchasers of its chips. Such arrangements can accelerate construction but also expose the company to customers whose revenues remain dependent on fast growth in AI demand.
OpenAI has sought to secure computing capacity through agreements with Microsoft, Oracle, Amazon and other infrastructure providers while pursuing greater control over facilities designed for training and running advanced models. The Ohio proposal would give it access to a scale of power and processing capacity far beyond that of a conventional corporate data centre.
The project also fits within the broader Stargate infrastructure strategy promoted by OpenAI and SoftBank. Announced with a goal of mobilising up to $500 billion for US AI infrastructure, Stargate is intended to combine capital, energy resources, semiconductor technology and data centre development across multiple sites.
Nvidia and OpenAI announced a separate strategic partnership in September 2025 that envisaged deploying at least 10 gigawatts of Nvidia systems. Nvidia said it intended to invest up to $100 billion progressively as each gigawatt entered service. The Ohio negotiations appear to involve a different financial structure, focused on guaranteeing obligations and supporting equipment purchases rather than taking ownership of the entire project.
The scale of the proposed arrangement has renewed scrutiny of circular financing across the AI sector. Nvidia could help finance a customer that would use the money to lease facilities filled largely with Nvidia equipment, supporting both chip sales and the customer’s ability to consume computing services.
Supporters of the model argue that supplier-backed financing is common in capital-intensive industries, including aviation, energy and telecommunications. Guarantees can help establish new markets when customers face exceptionally high upfront costs and lenders lack experience assessing the underlying technology.
Critics warn that the structure may transfer significant risk to Nvidia if OpenAI’s revenue growth fails to match its infrastructure commitments. Large language-model developers are spending heavily on computing, electricity and engineering while facing pressure to turn rapid adoption into durable profits.
Global investment in data centres, processors and supporting power systems is accelerating as technology groups race to build larger models and provide AI services to businesses and consumers. Projects are increasingly constrained by access to electricity, grid connections, construction labour, cooling systems and specialised components rather than by demand for land alone.
Southern Ohio offers access to extensive industrial sites and proximity to energy infrastructure, but a 10-gigawatt campus would require years of approvals and coordinated investment. The development is expected to use federal and privately controlled land, while its power plan could include a $33 billion natural gas facility supported through economic cooperation between the United States and Japan.
Follow Arabian Post
Select Arabian Post as your preferred source on Google and MSN News for trusted business news and Arab politics and updates.
