SpaceX announced Tuesday that it is partnering with Nvidia to design the compute payload for Starmind AI1, the first satellite in a planned constellation built to run AI workloads directly in orbit, according to a post on X.
What Happened
Each Starmind satellite will carry Nvidia’s Rubin GPUs and Vera CPUs, per SpaceX’s announcement, which included renderings of the payload design. The news landed hours before SpaceX’s first earnings call as a public company, where Musk said the company has committed to building its AI infrastructure exclusively on Nvidia hardware.
“We think the Vera Rubin architecture is the best architecture. We think it’s the best AI computer, and we greatly value our close cooperation and partnership on many levels with Nvidia,” Musk told investors. “So we’re exclusive to Nvidia.”
Musk said SpaceX plans to deploy Nvidia’s Vera Rubin NVL72 rackscale system, codenamed Kyber, both on the ground and in space, targeting 2 gigawatts of compute capacity online by the end of this year and roughly 10 gigawatts by the end of 2027.
Starmind has been in development since Musk confirmed the name in June, following an xAI trademark filing. The concept involves satellites with onboard processors and large solar arrays computing AI workloads in orbit and beaming results back to Earth, rather than moving data down to ground-based servers. SpaceX has filed with the FCC for a constellation of up to one million satellites, citing constant solar power and the absence of zoning restrictions as advantages over terrestrial data centers.
Why It Matters
The Nvidia exclusivity marks a shift in tone from two weeks ago, when Musk dismissed a report that SpaceX had ordered $52 billion worth of Nvidia GPUs through Foxconn as fake news. While the dollar figure may have been wrong, the underlying direction appears correct: SpaceX’s AI division already leases Colossus compute capacity to Anthropic and Google, and Tuesday’s earnings report showed AI revenue climbing sharply as those deals ramp up.
Nvidia shares rose roughly 3% in Tuesday trading on the news, while SpaceX stock climbed nearly 9% during the day before giving back gains after hours as investors digested the earnings report’s capital spending figures. The partnership positions Nvidia at the center of an orbital computing push that could reshape how AI workloads are deployed, though the constellation remains early-stage and subject to regulatory approval.
The Bottom Line
SpaceX and Nvidia have formalized a hardware alliance for Starmind, with Rubin GPUs and Vera CPUs heading into orbit as part of Musk’s broader AI infrastructure plan. Targets call for 2 gigawatts of compute online by year-end and 10 gigawatts by 2027, but the program still faces FCC review and scaling challenges.