Inside the $4 billion push to build brand-new energy generation specifically for a single AI hub
When SoftBank and SB Energy committed more than $4 billion to build at least 10 gigawatts of brand-new energy generation for a single data center in Ohio, they weren't just solving a power problem.
They were signaling a fundamental shift in how the AI industry thinks about electricity — treating it not as a utility cost, but as the core bottleneck for growth.
The money will go directly into local energy infrastructure, including new power plants or renewable projects, all dedicated to feeding one massive AI hub.
That hub, leased by OpenAI for 20 years and backed by a $105 billion financing commitment from Nvidia, will sit on the site of a former Cold War-era uranium enrichment plant.
It could eventually consume up to 8 gigawatts — more electricity than the entire city of San Francisco used in 2024, which clocked in at just over 5 GW.
To put that in perspective, a typical nuclear power plant produces around 1 GW. This single data center could demand the output of eight such plants.
And OpenAI, SB Energy, and Nvidia are building that capacity from scratch, not relying on existing grids that are already straining under rising demand from other data centers, factories, and homes.
Our analysis suggests this approach reflects a growing recognition that the grid cannot keep up with AI's appetite.
Rather than waiting for utilities to expand — a process that can take a decade — companies are taking energy production into their own hands.
The $4 billion figure, while large, is only part of the story.
Nvidia is also investing $1.5 billion directly into SB Energy, and OpenAI is putting $40 million toward "local priorities," likely to ease community concerns.
Those concerns are real. More than 70 percent of Americans oppose building data centers near their homes, according to a Gallup poll from March.
Rising electricity bills, partly blamed on these facilities, are becoming a midterm election issue.
OpenAI and SB Energy insist that the data center's energy and infrastructure costs will not be passed on to local residents — a promise that will be tested as construction begins.
The project is expected to create 35,000 temporary construction jobs over six years and 2,500 long-term operating jobs. But the energy investment itself may have a longer shadow.
If successful, this model — building dedicated power generation for a single AI hub — could become the template for future mega-projects.
It would mean that the next wave of AI infrastructure is not just about chips and cooling, but about owning the power supply.
OpenAI CEO Sam Altman called it "a huge site, with enough computing power to help millions of people use AI to do things we can only start to imagine today." But the energy behind that imagination is being built brick by brick, gigawatt by gigawatt, in rural Ohio.
Original reporting sourced from BSS/AFP, Nvidia, OpenAI, SoftBank.
