NVIDIA is investing $1.5 billion in SoftBank-backed SB Energy as part of a major artificial intelligence infrastructure project in Pike County, Ohio, where SB Energy plans to build a data-center campus for OpenAI.
The facility is expected to reach 8 gigawatts of IT capacity, making it one of the larger AI-computing projects announced in the United States.
The investment is only one part of a much larger arrangement. NVIDIA will provide credit support tied to land, power and the data-center construction, while becoming the exclusive AI compute infrastructure provider at the campus. OpenAI will be the customer and has agreed to a 20-year lease. Initial capacity is planned at 4.25 gigawatts, with an option for another 3.75 gigawatts.
The project is important because AI infrastructure is increasingly constrained not only by chips, but by the availability of electricity, land, grid connections and financing. The Ohio development brings all four into the same deal.
NVIDIA puts $1.5 billion into SB Energy
NVIDIA says the capital will support SB Energy’s growth and its commitments to the Ohio community. SB Energy is already backed by SoftBank Group and OpenAI, which invested $1 billion in the company earlier this year.
The chipmaker says it will secure the site’s land, power and shell capacity, with the campus designed specifically to host NVIDIA AI systems. The project will use NVIDIA’s DSX AI factory platform, including its GPUs, CPUs and networking equipment.
The Ohio AI data center will be built for OpenAI
SB Energy will build, own and operate the data-center. OpenAI will lease the facility for 20 years and use the computing capacity for its AI workloads.
NVIDIA has an option to extend the arrangement by another 3.75 IT gigawatts, bringing the planned total to 8 IT gigawatts. The first capacity is expected to come online in phases beginning in 2028.
[ALSO READ: Intel Invests $28 Billion for two new chip factories in Ohio ]
Why power has become as important as GPUs
The Ohio project shows how the AI data-center business is changing. A few years ago, the main discussion around AI infrastructure was about access to GPUs. Today, a developer can have demand from customers and still struggle to build a data center because it cannot secure enough electricity or grid capacity.
SB Energy and SoftBank plan to develop at least 10 GW of new power generation for the project and invest at least $4.2 billion in regional grid infrastructure through a partnership with AEP Ohio.
The project is also being developed on and around the site of the former Portsmouth Gaseous Diffusion Plant, bringing a large new industrial use to a site with an existing history of energy-intensive operations. SB Energy says the campus spans private and federal land and is being developed with AEP Ohio, the U.S. Department of Energy and the U.S. Department of Commerce.
OpenAI’s role is becoming much larger
OpenAI and SoftBank announced their $1 billion investment in SB Energy in January 2026, with OpenAI initially signing a 1.2-GW data-center lease.
The latest agreement is far larger, with the Ohio campus now planned for as much as 8 GW.
For OpenAI, long-term access to dedicated infrastructure reduces one of the biggest uncertainties around operating large AI systems: whether enough compute will be available when it is needed. It also gives OpenAI a closer relationship with the physical infrastructure supporting its models.
What it means for the AI infrastructure market
The bottleneck is no longer simply how many GPUs can NVIDIA manufacture. It is increasingly about whether companies can bring together: chips + power + land + data centers + grid connections + financing.
NVIDIA’s $1.5 billion investment in SB Energy gives the company a position on several of those layers.
It also shows a wider move by NVIDIA into infrastructure financing. The company announced earlier in August that it was working with major financial institutions on AI-compute infrastructure financing platforms intended to mobilize more than $500 billion in third-party capital.
[ALSO READ: L&T Wins Up to ₹10,000–15,000 Crore Order to Build India’s Largest Single-Cluster AI Factory With 10,000 NVIDIA B300 GPUs ]
The important point about the Ohio project is that NVIDIA will no longer act only like a semiconductor vendor.
AI infrastructure has become sufficiently capital-intensive that the boundaries between chip supplier, customer, data-center operator and financier are becoming less clear.
The company is effectively helping secure the conditions under which its own computing systems can be deployed. That is strategically rational when demand is strong and data-center capacity is scarce.
But the model also increases the industry’s financial interdependence. NVIDIA needs customers such as OpenAI to build profitable AI businesses; OpenAI needs huge amounts of computing infrastructure; infrastructure developers need financing and long-term tenants; and NVIDIA is increasingly participating in that financing.
