Nvidia says that it plans to invest $100 billion in OpenAI to partner on building new data centers and artificial intelligence infrastructure.

According to a joint release from the two companies on September 22, OpenAI and Nvidia have signed a letter of intent, but are still working to finalize the deal in the coming weeks. Under the current version of the agreement, OpenAI will look to build and deploy data centers requiring a combined 10 gigawatts of power capacity, the equivalent of enough energy to power more than 8 million homes in a single year. Servers in the data centers will be equipped with Nvidia's advanced chips to train and operate AI models. 

“Everything starts with compute,” said OpenAI CEO Sam Altman. “Compute infrastructure will be the basis for the economy of the future, and we will utilize what we’re building with NVIDIA to both create new AI breakthroughs and empower people and businesses with them at scale."

CNBC reports that Nvidia's investments in OpenAI will come in phases, starting with $10 billion when the agreement is made final and the first gigawatt of computing power comes online, while subsequent investments will come in as additional gigawatts are spun up. Nvidia will also receive equity in OpenAI, and be OpenAI's preferred supplier for chips and networking equipment. The companies are hoping to complete the first phase of the project sometime in the back half of 2026, but have not yet specified where exactly the data centers will be located. 

This comes days after Nvidia announced that it would be investing $5 billion in its struggling competitor Intel, in exchange for Intel agreeing to integrate Nvidia's technology into its AI infrastructure platforms. In late August, the Trump administration obtained a 10% stake in Intel, as part of a bid from the White House to boost domestic chip manufacturing and compete with China's growing dominance in the semiconductor sector.