Quick Answer
Nvidia has agreed to guarantee up to $105 billion for OpenAI to lease an Ohio data center that SB Energy will build, own, and operate. The proposed campus starts at 4.25 gigawatts and could reach about 8 gigawatts, while financing remains unfinished. The practical point is that the arrangement shifts a large share of infrastructure risk into a long-term Nvidia commitment.
Key Takeaways
- Nvidia disclosed a guarantee of up to $105 billion for OpenAI’s Ohio data center lease on August 17, 2026.
- SB Energy plans to build, own, and operate the PORTS-Pike Technology Campus in Pike County, Ohio.
- The project begins with 4.25 GW of computing capacity and includes an option for another 3.75 GW.
- OpenAI is expected to serve as the anchor tenant under a 20-year lease, with initial capacity targeted for 2028.
- The financing structure still requires debt and equity arrangements, which means the final project economics are not yet settled.
What did Nvidia agree to guarantee for OpenAI’s Ohio data center?
Nvidia agreed to provide a guarantee of up to $105 billion to help OpenAI lease a large data center in Ohio, according to a securities filing disclosed on August 17, 2026. The agreement concerns the PORTS-Pike Technology Campus, which SoftBank-owned SB Energy plans to build in Pike County, Ohio. Nvidia also agreed to make a separate $1.5 billion direct equity investment in SB Energy.
The Nvidia guarantee is significant because OpenAI’s planned computing expansion requires more than chips alone. A project of this scale needs land, power connections, buildings, cooling systems, networking equipment, financing, and a long-term customer willing to pay for capacity. CNBC reported that OpenAI will be the anchor tenant under a 20-year lease, while SB Energy will build, own, and operate the facility.
Nvidia’s commitment does not mean Nvidia is writing OpenAI a $105 billion cash payment. A guarantee generally supports a borrower’s or tenant’s ability to secure financing by assuring lenders or counterparties that obligations will be met under specified conditions. The final terms matter because the filing disclosed the maximum guarantee amount, while the full debt and equity structure remains incomplete.
How large is the planned OpenAI facility in Pike County, Ohio?
The Ohio data center is planned to start with 4.25 gigawatts of computing capacity and could expand by another 3.75 gigawatts, for total planned capacity of about 8 gigawatts. That scale would place the project among the largest AI infrastructure proposals disclosed publicly, because gigawatt-level facilities require substantial power generation, transmission, cooling, and physical construction.
The PORTS-Pike Technology Campus is located on the former Portsmouth Gaseous Diffusion Plant, a 3,777-acre former uranium enrichment facility operated by the U.S. Department of Energy. The location gives the project a large industrial footprint, although a suitable site does not remove the need for extensive power and construction planning.
| Project element | Disclosed detail | Why it matters |
|---|---|---|
| Initial capacity | 4.25 GW | The first planned phase establishes the minimum scale of the proposed computing campus. |
| Expansion option | 3.75 GW | The option could bring the total planned capacity to about 8 GW. |
| Initial capacity target | About 800 MW in 2028 | The first operational phase is expected before the full campus is complete. |
| Potential full buildout | 2032 | The project remains a multi-year construction and financing effort. |
| Anchor tenant | OpenAI under a 20-year lease | The lease provides the long-term demand case for the project. |
Data Center Knowledge reported that the site’s first capacity, roughly 800 megawatts, is expected to come online in 2028. Full buildout could extend to 2032, which means the stated 8 GW figure is a long-term target rather than capacity OpenAI can use immediately.
How will Nvidia and SB Energy finance the Ohio data center?
The Ohio data center financing is not finalized, and the project is expected to use a mixture of debt and equity. Reuters, through Euronext’s financial news service, reported that possible funding sources include project finance loans or bonds, direct SoftBank investment, and potential equity from a future SB Energy initial public offering.
The financing structure matters because a data center of this size requires capital long before it can generate lease revenue. Project finance typically relies on expected future cash flows and long-term contracts, which makes OpenAI’s role as anchor tenant central to the investment case. Nvidia’s guarantee can help reduce lender concerns, but the guarantee also links Nvidia more closely to the financial outcome of a major customer’s infrastructure plan.
SB Energy’s ownership role creates a division of responsibilities. SB Energy is expected to construct and operate the campus, while OpenAI is expected to lease the computing capacity and Nvidia is expected to supply the AI computing platform. The practical limitation is that each part of the arrangement depends on the others: construction, power availability, financing, customer demand, and hardware deployment all need to align over several years.
Why is Nvidia backing OpenAI’s infrastructure so directly?
Nvidia is backing the Ohio project because the company wants to secure long-lived infrastructure for its computing systems, according to Nvidia CEO Jensen Huang. Huang told CNBC that Nvidia is securing infrastructure so OpenAI can deploy AI factories that can be upgraded repeatedly. Nvidia is expected to be the exclusive AI compute and chip provider for the site, supplying GPUs, CPUs, networking equipment, and infrastructure software through its DSX AI factory platform.
The arrangement gives Nvidia a path to place large volumes of its hardware into an OpenAI-focused facility over time. Huang said each hardware generation deployed at the site could represent roughly 1.5 million Nvidia GPUs, a figure that could translate to about $150 billion to $200 billion in Nvidia revenue per generation through 2030, according to Yahoo Finance’s report. Those figures reflect Huang’s projections and should be treated as company estimates rather than independently verified future revenue.
OpenAI also benefits because guaranteed access to large amounts of compute could support future model training and AI services. The company’s computing requirements are becoming a central business issue as its products expand, a pressure that also shapes discussion around OpenAI’s revenue growth and IPO expectations. The limitation is that access to infrastructure does not by itself establish how much revenue OpenAI can generate from the resulting AI products.
Why are investors examining Nvidia’s AI financing commitments?
Nvidia’s Ohio commitment is drawing scrutiny because Nvidia sells the computing hardware that OpenAI is expected to use at the facility. Critics of similar arrangements question whether chip makers, infrastructure companies, lenders, and AI developers can become too financially dependent on one another. The concern is not that the Ohio agreement is automatically improper, but that overlapping commitments can make it harder to separate customer demand from vendor-supported expansion.
Huang said the arrangement is “not circular financing” and argued that Nvidia is using its scale and long-term visibility to help support AI infrastructure. The South China Morning Post reported that the announcement followed Nvidia’s financing partnerships with six major financial institutions, including BlackRock, targeting more than $500 billion in third-party AI infrastructure funding.
The Ohio project is separate from Nvidia’s earlier AI financing platform, which focused on third-party infrastructure funding rather than one named OpenAI facility. The practical distinction matters because the $105 billion guarantee is tied to a specific Ohio lease and construction plan. Investors and customers should still recognize that final financing terms, construction schedules, and capacity demand can change before the full campus is completed.
What does the Ohio project mean for OpenAI’s AI capacity?
The Ohio project could give OpenAI access to a major long-term source of computing capacity, beginning with roughly 800 megawatts expected in 2028. OpenAI President Greg Brockman described compute as a limited resource for the AI industry, comparing its strategic importance to oil. The statement reflects OpenAI’s view that access to data center capacity is becoming as important as software development for leading AI companies.
Large AI models require extensive computing resources for training and operation, including specialized processors, high-speed networking, power delivery, and cooling. A dedicated campus can make capacity planning more predictable than relying only on short-term cloud availability. At the same time, the project’s long schedule means OpenAI cannot treat its eventual 8 GW target as immediate capacity for current products.
For consumers, the deal does not create an immediate change to ChatGPT pricing, feature access, or model availability. The more direct effect is on the industry’s ability to build larger AI systems over time. Users evaluating new AI products should continue to consider privacy, reliability, and pricing separately from the size of the data center behind the service, especially as companies introduce products such as teen-focused AI safety features.
What happens next for Nvidia’s OpenAI Ohio data center agreement?
The next stage for Nvidia’s OpenAI Ohio data center agreement is the completion of the financing structure and the start of construction planning. The reported structure includes potential debt through project finance loans or bonds and possible equity sources involving SoftBank or SB Energy. Those arrangements will determine how the $105 billion guarantee operates in practice and what obligations each party accepts.
The first major operational milestone is expected in 2028, when approximately 800 megawatts of capacity could come online. The larger 4.25 GW initial configuration and optional expansion to about 8 GW would take much longer, with full buildout potentially reaching 2032. Construction timing, electricity availability, financing conditions, and demand for AI computing will all affect whether the project follows that schedule.
The most sensible interpretation is that Nvidia is making a substantial strategic infrastructure commitment rather than announcing a completed 8 GW data center. OpenAI, Nvidia, SB Energy, lenders, and power providers still need to execute a complex multi-year plan. Readers should watch for finalized financing documents, construction updates, power agreements, and more detailed disclosures about the guarantee’s terms.
FAQ
Is Nvidia giving OpenAI $105 billion in cash?
Nvidia is not reported to be giving OpenAI $105 billion in cash. Nvidia agreed to provide a guarantee of up to $105 billion to support OpenAI’s lease of the Ohio data center, while the final financing structure remains unfinished.
Where will OpenAI’s Ohio data center be located?
OpenAI’s planned Ohio data center will be located at the PORTS-Pike Technology Campus in Pike County, Ohio. The campus is on the site of the former Portsmouth Gaseous Diffusion Plant, a former uranium enrichment facility.
How much computing capacity will the Ohio data center have?
The Ohio data center is planned to begin with 4.25 GW of computing capacity and could add another 3.75 GW. The potential total capacity is about 8 GW, although the full buildout could take until 2032.
When will the OpenAI Ohio data center begin operating?
The OpenAI Ohio data center is expected to bring roughly 800 megawatts of initial capacity online in 2028. The project’s broader capacity expansion is expected to continue over several years after that first phase.
Why is Nvidia’s OpenAI financing deal receiving scrutiny?
Nvidia’s OpenAI financing deal is receiving scrutiny because Nvidia is both backing infrastructure financing and supplying the AI hardware expected to be used at the site. Nvidia CEO Jensen Huang has rejected the characterization of the arrangement as circular financing, but final financing details will remain important for investors to assess.