Microsoft and OpenAI are working on a new US-based AI supercomputer data center. Called Stargate because such a futuristic-sounding project deserves an equally sci-fi-inspired name, it is planned to power OpenAI's next generation of ChatGPT-like AI systems and could launch as early as 2028.

Stargate is estimated to be one of the largest and most advanced data centers in the world, spanning hundreds of acres and using up to 5 gigawatts of power. Of the series of facilities the two companies plan to build over the next six years, Stargate will be the largest and is considered crucial to Open AI's ability to train and operate more advanced AI models than ChatGPT-4. Given its power needs, both companies are considering the use of alternative power sources such as nuclear energy.

The name "Stargate" comes from the 1994 science fiction film of the same name, in which an archaeologist discovers an interstellar teleportation device that transports him to an alien world that has had a hand in human civilization. The film's trailer famously describes it as "a key to the past, a doorway to the future, a passageway to discovery," which seems to be the obvious inspiration for OpenAI and Microsoft's AI supercomputer venture.

OpenAI CEO Sam Altman and Microsoft employees have divided their plan to build multiple supercomputers across the United States into five phases, with Stargate reportedly the fifth and final phase, according to The Information, Stargate supercomputer project is estimated to take five to six years to complete. Microsoft also aims to launch a smaller, fourth stage supercomputer for OpenAI around 2026.

To date, Microsoft has poured more than $13 billion into OpenAI for startups that use Microsoft data centers to power ChatGPT. But Stargate will be an unprecedented (and expensive) expansion of that partnership. According to sources at The Information who viewed Microsoft's initial cost estimate and spoke with Altman about the proposal, the project could cost as much as $100 billion. This staggering figure is more than 100 times the cost of the largest data center currently in operation and more than three times what Microsoft spent last year on servers, buildings, and other equipment.

It is important to note that the project has not yet been approved and it is unclear where in the U.S. it could be headquartered if it moves forward. Of course, it all depends on OpenAI's ability to deliver the next-generation large-scale language model rumored to be GPT-5. Last year, the company was scheduled to present a new AI project to Microsoft called Arrakis (another science fiction reference, this time "Dune"), but missed the opportunity, citing the limitations of current supercomputers as a bottleneck preventing development.

Stargate may still be years away, but many AI-powered tools are already available. Microsoft has made OpenAI's largest language model available for free on its Copilot platform. OpenAI also recently made ChatGPT available to anyone without an account.