Key Points:
- Microsoft and OpenAI collaborate on a data center project that could cost up to $100 billion. They aim to launch the “Stargate” AI supercomputer by 2028.
- The project reflects the surging demand for AI data centers capable of handling complex tasks driven by generative AI technology.
- Microsoft is expected to finance the project, which could be a hundred times costlier than existing data centers.
- The project is organized into five phases, with “Stargate” as the fifth phase, following a smaller supercomputer slated for launch in 2026.
Microsoft and OpenAI are reportedly joining forces on an ambitious data center project estimated to cost as much as $100 billion. The project aims to launch an artificial intelligence supercomputer, “Stargate,” by 2028. It underscores the escalating demand for AI data centers capable of handling increasingly complex tasks driven by the rapid adoption of generative artificial intelligence technology.
According to The Information, Microsoft is likely to spearhead the project’s financing, which could be a hundred times costlier than existing data centers, citing insights from individuals familiar with private discussions. The proposed supercomputer, slated for construction in the U.S., is envisioned to be the largest in a series of supercomputers planned for development over the next six years.
Sources familiar with the project’s details, including discussions with OpenAI’s CEO, Sam Altman, and Microsoft’s initial cost assessments, attributed the estimated $100 billion cost. The project has five phases, with “Stargate” as the fifth. Preceding this, Microsoft is already working on a fourth-phase supercomputer dedicated to OpenAI, which is slated for launch around 2026.
The substantial cost projections for the upcoming phases primarily involve the acquisition of specialized AI chips, which are known for their high prices. Nvidia’s CEO, Jensen Huang, recently revealed that the latest “Blackwell” B200 AI chip would be priced between $30,000 and $40,000. Microsoft announced the development of custom-designed computing chips in November of last year.
The project is designed to be compatible with chips sourced from various suppliers, reflecting a strategic approach to harnessing diverse technological capabilities in AI development. A Microsoft spokesperson acknowledged the company’s continuous focus on infrastructure innovations to advance AI capabilities without directly confirming details about the Stargate supercomputer’s planned launch.
The anticipated expenses for the project could surpass $115 billion, marking a significant increase compared to Microsoft’s capital spending on servers, buildings, and equipment in the previous year. The collaboration between Microsoft and OpenAI underscores their shared commitment to driving innovation in artificial intelligence and positioning themselves at the forefront of technological advancement.
