OpenAI has taken another major step toward expanding its artificial intelligence infrastructure by securing a 20-year lease for a massive data center campus in Ohio. The agreement covers an initial 4.25 gigawatts of AI computing capacity, with Nvidia providing financial guarantees of up to $105 billion.
The project will be developed at the PORTS-Pike Technology Campus in Pike County, Ohio, with the first sections expected to become operational in phases beginning in 2028.
The scale of the agreement highlights just how quickly demand for AI computing power, electricity, and data center infrastructure is growing.
OpenAI Plans Massive AI Expansion in Ohio
Under the long-term agreement, OpenAI will use the PORTS-Pike campus to support its growing AI computing requirements.
SB Energy will be responsible for building, owning, and operating the data center infrastructure, while Nvidia will serve as the campus’s exclusive AI computing technology provider.
The initial development will provide approximately 4.25 GW of IT capacity. Nvidia has also secured an option for an additional 3.75 GW, which could eventually increase the campus’s total capacity to around 8 GW.
If fully developed, the project would become one of the largest AI infrastructure developments in the United States.
The campus will be built on the site of the former Portsmouth Gaseous Diffusion Plant and surrounding land in southern Ohio. Development of the project involves cooperation with AEP Ohio, the U.S. Department of Energy, and the Department of Commerce.
Nvidia Backs the Project With Guarantees of Up to $105 Billion
One of the most significant parts of the agreement is Nvidia’s financial backing.
According to Nvidia’s regulatory filing, the company has entered into several residual-value guarantee agreements connected to OpenAI’s initial 4.25 GW commitment. Nvidia’s total potential payment obligations under these agreements are capped at $105 billion.
However, this does not mean Nvidia will directly pay OpenAI’s rent.
OpenAI remains responsible for making payments under its lease. Nvidia’s guarantee would generally become relevant only under specific circumstances, such as OpenAI becoming insolvent and failing to meet its lease obligations.
In that situation, Nvidia could be responsible for covering part of the difference between the guaranteed minimum value of the lease and the amount recovered by SB Energy through another tenant or a sale of the property.
Nvidia could also have several alternatives, including taking over the lease, helping find another tenant, or initiating a sale process.
OpenAI has agreed to reimburse Nvidia for payments the chipmaker may have to make under these guarantees.
Depending on the individual section of the campus, Nvidia’s guarantee could remain in place for as long as the 20-year lease period, although certain conditions could allow the obligations to end earlier.
Nvidia Says OpenAI Will Make the Lease Payments
The unusual size of Nvidia’s guarantee has raised questions about the financial relationship between major companies involved in the AI infrastructure boom.
Nvidia CEO Jensen Huang has emphasized that OpenAI, not Nvidia, will be responsible for the lease payments.
He also argued that Nvidia’s computing infrastructure is widely usable across the AI industry. If OpenAI were unable to use the capacity, the infrastructure could potentially be offered to other cloud providers, enterprises, AI laboratories, or startups.
That flexibility is particularly important for a project involving billions of dollars of infrastructure designed to operate for decades.
OpenAI Will Use Nvidia’s AI Technology Stack
The Ohio campus will rely heavily on Nvidia technology.
OpenAI is expected to deploy Nvidia’s DSX AI factory platform across the initial 4.25 GW development, subject to certain exceptions contained in the agreements.
The infrastructure will combine Nvidia GPUs, CPUs, networking technologies, software, and supporting data center systems.
Modern AI models require enormous amounts of computing power, but obtaining GPUs is only one part of the challenge. AI companies increasingly need access to large amounts of electricity, suitable land, cooling infrastructure, transmission capacity, and buildings capable of supporting high-density computing systems.
As AI development accelerates, access to power and data center capacity is becoming nearly as important as access to advanced processors.
Nvidia Will Invest $1.5 Billion in SB Energy
Nvidia’s involvement goes beyond providing computing equipment and financial guarantees.
The chipmaker also plans to invest $1.5 billion in SB Energy, joining investors that include SoftBank Group and OpenAI.
The investment is expected to support SB Energy’s growing portfolio of data center and energy infrastructure projects.
Meanwhile, SB Energy and SoftBank plan to develop at least 10 GW of new electricity generation to support approximately 8 GW of IT capacity.
The companies are also expected to invest at least $4.2 billion in regional grid infrastructure through their arrangement with AEP Ohio.
A key goal of the energy agreement is to ensure that the enormous infrastructure requirements of the AI campus do not shift the project’s development costs onto existing Ohio electricity customers.
Thousands of Jobs Expected From the Ohio AI Campus
The project could also have a major economic impact on southern Ohio.
The development is expected to support approximately 35,000 construction jobs through 2032 and create around 2,500 permanent operational positions once the campus is running at scale.
OpenAI and SB Energy have also established an $80 million community benefits fund.
The funding is intended to support initiatives including workforce training, affordable energy programs, employment opportunities, and broader economic development in the surrounding communities.
OpenAI CEO Sam Altman has said the company wants local residents to benefit from the project through employment, business opportunities, and community investment.
AI Companies Are Competing for Power and Data Center Land
OpenAI’s Ohio agreement reflects a much broader change taking place across the technology industry.
As AI models become larger and more widely deployed, technology companies need increasingly powerful computing infrastructure. Building that infrastructure requires more than purchasing GPUs. Companies must secure electricity generation, grid connections, land, cooling systems, and large-scale data center facilities.
Many of these resources can take years to develop.
This growing demand has even increased interest in infrastructure originally developed for cryptocurrency mining. Bitcoin mining companies often already control valuable assets such as powered land, grid connections, substations, and large computing facilities.
Some miners have consequently started converting existing or planned sites into high-performance computing and AI data centers.
OpenAI Continues to Diversify Its Computing Infrastructure
The Ohio project is also part of OpenAI’s broader effort to secure enough computing capacity for its future AI systems.
The company has been expanding its infrastructure relationships rather than depending entirely on a single cloud or technology provider.
This strategy gives OpenAI access to more computing resources as demand for AI products and increasingly sophisticated models continues to rise.
At the same time, reports have continued to speculate about a possible OpenAI public listing. A potential IPO has been discussed for 2027, although no public listing has been officially confirmed.
What the OpenAI-Nvidia Ohio Deal Means for the AI Industry
The most important part of the agreement may not simply be its enormous dollar value.
The deal demonstrates how the AI industry’s competition is increasingly moving beyond individual GPUs and models toward energy, land, data centers, and long-term computing capacity.
A 20-year commitment for 4.25 GW of initial AI capacity shows that OpenAI is planning its infrastructure requirements far into the future. The possibility of expanding the campus to 8 GW makes the project even more significant.
Nvidia’s involvement also shows how the chipmaker is expanding its role beyond selling GPUs. Through hardware, networking, software, investment, and financial guarantees, Nvidia is becoming increasingly involved in the infrastructure required to build enormous AI computing facilities.
With the first capacity expected to arrive in 2028, the PORTS-Pike Technology Campus could become an important part of the next generation of U.S. AI infrastructure.
