Nvidia Confirms It Will Back Massive 4.25GW SoftBank Data Center In Ohio For $105 Billion
Nvidia has confirmed reports that it is financially backing SoftBank's massive data center at the Portsmouth Site in Pike County, Ohio, with up to $105 billion in financing, a securities filing revealed on Monday.
The project, led by SB Energy, is set to be fully leased to OpenAI. The facility could grow to 10GW, making it the world's largest data center, but Nvidia will initially support the first 4.25GW, DataCenterDynamics reported.
In a LinkedIn post, Nvidia CEO and founder Jensen Huang confirmed that his company would partner with SB Energy to help secure land, power, and shell (LPS) capacity at the Department of Energy (DOE) site.
The GPU giant will provide a $105bn backstop for the project, helping lower debt costs, similar to previous SPV project financings and reusing the structure popularized by Meta in late 2025 with its Project Beignet off balance sheet structure to fund its massive Hyperion data center . It had initially planned to support as much as $250bn, but faced investor pushback over the extent of the risk.
As a reminder, META is already neck deep in off-balance sheet debt. Here is a schematic of its $27.3 billion SPV with Blue Owl "Project Beignet" for the Hyperion data center. None of this touches META's balance sheet.
— zerohedge (@zerohedge) January 29, 2026
Expect hundreds of billions of these in 2026 https://t.co/794EgSiiZ9 pic.twitter.com/7hMyVW6Lno
Nvidia will also invest $1.5bn in SB Energy, down from a reported $3bn. The SoftBank subsidiary, which is preparing for an IPO, is set to build 10GW of new power generation, including 9.2GW of natural gas generation, to power the facility. This will provide up to 8GW of total IT load.
SB Energy and SoftBank also plan to invest at least $4.2bn in new regional grid infrastructure through a partnership with AEP Ohio they say is designed to protect ratepayers.
"Nvidia is supporting the LPS infrastructure at Ports-Pike for approximately 4GW over a 20-year term, securing a site on which Nvidia compute will be exclusively deployed," Huang said.
"Our support is limited to defined portions of lease and power payments, along with a specified residual-value commitment — not the full cost of the site or all of the tenant’s obligations. The guarantee will become effective in phases as data centers are placed in service between 2028 and 2030. As OpenAI makes lease payments and capacity comes online, Nvidia's remaining exposure declines."
Huang said that most companies will secure LPS independently, and then buy Nvidia gear, but that leading AI labs are "growing faster than their balance sheets and long-term credit profiles can support.
Ports-Pike will not only exclusively feature Nvidia GPUs, but also the company's full-stack DSX AI factory platform, including CPUs, networking, and infrastructure software. Over 20 years, it is expected to feature multiple generations of Nvidia gear.
"Each generation of Nvidia AI factory systems deployed at Ports-Pike could represent approximately 1.5 million Nvidia GPUs, or approximately $150 billion to $200 billion in Nvidia revenue," he said.
The company may yet decide to secure the remaining capacity of 3.75GW at the site.
Huang claimed that the deal was not an example of circular financing, an accusation the company has increasingly faced as it has funneled its profits back into the AI companies that buy its hardware.
"Nvidia uses its scale and long-term visibility to secure Ports-Pike to host Nvidia compute," he said. "This is the same discipline we apply to supply-chain management: we secure critical inputs when we have visibility into customer demand and when doing so enables long-term productive capacity."