Musk says SpaceX–Nvidia AI computer will reach volume production in 2027

Elon Musk says the AI computer SpaceX and Nvidia are developing together will reach volume production in 2027. His October 15 disclosure puts a manufacturing target on Star Mind, a joint project tied to the companies’ effort to design computing systems for data centers both on Earth and in space.

A server designed for two environments

Nvidia chief executive Jensen Huang described the engineering work during an appearance on the BG2 Pod. He said teams from both companies are developing a new server design and computing architecture intended to work in terrestrial and orbital settings. The collaboration therefore covers the design of the computing system, not simply the supply of Nvidia chips.

Musk shared a video of Huang discussing the partnership in his October 15 X post. He called the Star Mind computer “far better than almost anyone expects” and praised Nvidia’s engineers. That assessment is Musk’s characterization of the project; the concrete milestone he attached to it was the production year.

One particular new market that we're going to go explore together is putting data centers out in space

Jensen Huang, speaking on the BG2 Pod

Production is not deployment

The production target and the launch schedule describe different milestones. Musk has said initial satellite launches could begin in the fourth quarter of 2027, with significant deployment planned for 2028. Those remain prospective dates, rather than a network already being placed into service.

Benzinga’s account, published by Yahoo Finance, says the project involves adapting Nvidia’s Vera Rubin NVL72 computing system for space operations. It also describes an eventual network of up to one million satellites. That number is a proposed maximum constellation, not a firm commitment to manufacture or launch that many spacecraft.

The rocket is only part of the challenge

The orbital plan depends partly on Starship, SpaceX’s reusable rocket intended to make moving large amounts of hardware into space more economical. Benzinga cites Reuters reporting that the vehicle’s development has lagged behind Musk’s original targets. The computing project’s manufacturing schedule therefore sits alongside a separate launch-system dependency.

Launch economics are not the only obstacle. Benzinga identifies satellite maintenance, power supply and thermal management—controlling the computers’ heat—as major challenges for orbital data centers. The engineering task extends beyond getting servers into orbit to keeping them powered, maintained and operating at suitable temperatures.