It looks like OpenAI’s ChatGPT and Sora, among other projects, are about to get a lot more juice. According to a new report shared by The information, Microsoft and OpenAI are working on a new data center project, one part of which will be a massive artificial intelligence supercomputer called “Stargate.” Microsoft is said to be footing the bill, and the cost is as astronomical as the supercomputer’s name suggests — the entire project could cost upwards of $100 billion.

Spending more than $100 billion on anything is impressive, but when put into perspective, the price tag really shows how big an undertaking this could be: Reports claim that Microsoft and OpenAI’s new joint project could cost 100 times more than some of the largest data centers currently operating.

That estimate reportedly comes from people familiar with and involved in discussions about the project, including a person who spoke with Sam Altman, CEO of OpenAI, as well as someone with access to Microsoft’s initial cost analysis. The publication kept the sources anonymous.

The Stargate supercomputer is said to be based in the US and, as it reports Reuters, would be the largest of several facilities that Microsoft and OpenAI are preparing to create in the coming years. Stargate may not be released until 2028, but before it does, OpenAI will still receive periodic updates, with the next major AI update scheduled for early 2025.

Stargate marks the fifth and final (for now) phase of Microsoft and OpenAI’s supercomputing projects, and the companies are currently in the middle of phase three. Microsoft is reportedly already working on its fourth-phase supercomputer, which will be smaller than Stargate and could be launched by 2026.

It’s unclear what the Stargate might be capable of, but given the current AI boom and the need for AI models to keep evolving, it’s not surprising. GPT-5 is on the horizon, and for OpenAI and Microsoft to continue pushing the boundaries of AI, these massive data centers are a necessity.

As the joint venture could cost more than $115 billion when all is said and done, it’s worth noting that such an outlay would amount to more than three times what Microsoft spent on capital expenditures on buildings, equipment and servers during of 2023. Where will this money go? The Information claims that most of the cost of this five-phase plan comes from the need to purchase artificial intelligence chips. At this point, it’s a good time to be Nvidia.

Editors’ recommendations