SpaceX expects to launch its first artificial intelligence satellites powered by Nvidia Corp. technology in the fourth quarter of 2027.
Elon Musk says the orbital computing network will reach what he called “significant scale” by 2028, a faster timeline than previously announced.
When SpaceX went public earlier in June, it told investors that space launch tests would not happen until at least 2028.
Musk pulled that date forward during SpaceX’s first quarterly earnings report on August 4, 2026, surprising investors with the accelerated schedule.
SpaceX and Nvidia are working together to design the “compute payload” for Starmind AI1, the first in a planned network of orbiting AI satellites.
Each Starmind satellite will carry Nvidia’s Rubin GPUs and Vera CPUs, with Nvidia claiming its Space-1 Vera Rubin module delivers up to 25 times the AI processing performance of an H100 GPU.
On the earnings call, Musk announced that SpaceX has committed to building its AI infrastructure exclusively on Nvidia hardware going forward.
“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.
Musk added plainly: “So we’re exclusive to Nvidia,” underscoring the depth of the commercial relationship between the two companies.
He set a target of 2 gigawatts of compute capacity online by the end of this year, scaling to roughly 10 gigawatts by the end of 2027.
SpaceX envisions an orbital network that could eventually grow to one million satellites, creating a distributed AI supercomputer operating in space.
The constellation would rely on laser links to move data between satellites at extremely high speeds, bypassing traditional terrestrial data center limitations entirely.
Earlier this year, SpaceX asked the Federal Communications Commission for permission to deploy as many as one million orbital data center satellites between roughly 311 and 1,243 miles above Earth.
Musk described the satellite data center as “significantly simpler, lower cost, denser and lighter than a traditional rack” in a post on X.
Satellites operating in sun-synchronous orbit receive steady solar energy, giving SpaceX a consistent power source independent of Earth’s electrical grid.
Engineers also designed the Starmind architecture to remain hardware-flexible, allowing future satellites to swap processors without redesigning the entire spacecraft.
The partnership adds SpaceX to Nvidia’s growing list of space computing collaborators, after the chipmaker named six launch partners when introducing its space platform in March without including SpaceX.
