Nvidia has agreed to provide up to $105 billion in credit support for a massive new OpenAI data center campus in Ohio, according to a securities filing reported by the Financial Times. The financing covers the “land, power and shell” for the facility, which is being developed by SB Energy, a subsidiary of Japan's SoftBank, and will initially support 4.25 gigawatts of computing capacity with an option to expand by another 3.75 gigawatts.
The deal is notable as much for what it isn't as for what it is. Nvidia and OpenAI had reportedly been in talks for a backstop of up to $250 billion just weeks earlier, a figure that drew pushback from Nvidia investors concerned the chipmaker was putting an outsized share of its own balance sheet behind a bet designed largely to guarantee future demand for its GPUs. The final commitment came in at under half that number, with reporting from the Wall Street Journal describing it as a deliberate scale-back rather than a straightforward continuation of the earlier talks.
Financing Chips as Well as Buildings
The Ohio credit facility is only one piece of a larger financial relationship between the two companies. Nvidia is also discussing a separate arrangement to help finance OpenAI's chip purchases, a figure that reporting suggests could bring the total value of the combined project financing to as much as $350 billion. That structure, where a chip supplier helps underwrite its own customer's ability to buy its products, has become a recurring feature of the AI infrastructure buildout over the past year, and one that markets are increasingly parsing for signs of circularity risk rather than treating as a routine vendor-financing arrangement.
Why Markets Beyond AI Stocks Are Watching
Deals of this size move real capital through the broader economy, from power infrastructure to construction to regional labor markets, and they've become one of the clearer signals of how much conviction the biggest tech balance sheets still have in AI's growth trajectory even as other parts of the market show signs of strain. That question of conviction versus caution has been showing up elsewhere in macro data recently too, including in Japan's Q2 GDP growth slowing to just 1.1% against a 2.0% forecast, a reminder that not every part of the global economy is expanding at the pace AI infrastructure spending might suggest. Whether Nvidia's scaled-back commitment reflects genuine capital discipline or simply a renegotiated split of the same underlying bet will likely become clearer as the Ohio campus moves from financing terms to actual construction.
Related: Bitcoin ETF Outflows Top $390M as Fed Rate-Cut Bets Fade
