OpenAI and Oracle have announced the construction of 4.5 gigawatts of additional data centers in the United States as part of an expanded partnership that will help deliver on a massive plan to provide artificial intelligence capacity, Bloomberg reports.
OpenAI says the new facilities, along with a data center already under construction in Abilene, Texas, will provide more than 5 gigawatts of capacity. These data centers will house more than 2 million artificial intelligence chips.
The ChatGPT developer is not yet saying where the new facility will be built, but is considering the states of Texas, Michigan, Wisconsin, and Wyoming. OpenAI also notes that the new center will be built as part of the Stargate project, but this time without the involvement of one of the key investors, SoftBank.
As for the Abilene data center, OpenAI says the first building is already providing some of the company's compute workload for running and training algorithms. Oracle also began shipping the first racks of NVIDIA GB200 chips to the facility last month.
"We feel pretty good about our ability to move quickly on this because in many regards, Abilene was that beta test to prove out that you could build these at scale and at speed," said Chris Lehane, OpenAI's vice president of global policy.
Recall that at the end of 2024, xAI, an artificial intelligence startup owned by Elon Musk, announced its intention to launch a supercomputer for AI training, which will be equipped with a million chips.