Microsoft and OpenAI to build $100 billion supercomputer to power AI?

Microsoft and OpenAI are reportedly planning a supercomputer which will uses millions of specialized server chips and may cost up to $100 billion, The Information reported. The US-based supercomputer will be known as “Stargate” and will be the centre of a five-phase plan of the companies on a series of supercomputer installations in the next six years, the outlet reported. Stargate, could launch as soon as 2028, the outlet reported citing people in the know, who also said that executives at both companies have already drawn up plans for the data center project.
A spokesperson for Microsoft said as per the report, “We are always planning for the next generation of infrastructure innovations needed to continue pushing the frontier of AI capability.”
Microsoft has already committed more than $13 billion to OpenAI and is likely to fund Stargate as well, per the report. OpenAI currently uses Microsoft data centers to power its generative AI system ChatGPT for which Microsoft has exclusive rights to resell OpenAI’s technology to its own customers.
The supercomputer could be 100 times more expensive than the largest data centers currently in operation, it was reported. It could also have the potential to exceed computing power currently supplied by Microsoft to OpenAI from its data center. The supercomputer would require at least several gigawatts of power to do so, The Information reported.
The report also claimed that Stargate would be expensive owing to the difficulty in obtaining necessary server chips. Meanwhile, finding enough energy sources to power the project could also pose challenges and the companies have talked about using alternative power sources like nuclear energy, it added.
It’s not clear where Stargate would be located or whether it would be built in just one data center or many nearby centers, the report added.