SpaceX valuation rides on space data centers despite expert skepticism

Sam Altman, a tech executive who has criticized SpaceX’s space data center plans, traded barbed social media posts with Elon Musk, the leader of SpaceX, over the weekend. Responding to Musk accusing him of being a scammer, Altman wrote: “homeboy you’re the one sellling [sic] public market investors on short-term space datacenters.” The exchange put a spotlight on a broader disagreement over whether SpaceX’s orbital data center ambitions are close to becoming a real business, rather than just a talking point for public market investors.

SpaceX’s plans to launch a fleet of orbital data centers that perform AI inference — running data through trained models to produce outputs — are the main driver behind the company’s two-trillion-dollar valuation. But subject-matter experts, including entrepreneurs behind rival space data center startups and the team at Google developing its own orbital compute project, say space data centers are not going to be a serious business anytime soon, despite the enthusiasm reflected in SpaceX’s valuation.

Musk’s rebuttal centers on Starship, SpaceX’s rocket, which is expected to make its thirteenth test flight as soon as July 16. Even if the vehicle recovers both stages this time, operational reusable flight — recovering and reflying the same stages repeatedly — is likely still years away, a timeline that runs alongside SpaceX’s separate launch commitments to NASA, the government agency to which the company owes launch obligations. SpaceX also conceded during its IPO road show, the investor presentations companies hold before going public, that Starship may not be fully reusable in the near-term and will need to expend its second stages during each launch. Musk has said “we start flying them next year,” but the larger question is whether SpaceX can launch and manufacture such satellites at scale — likely a question for the 2030s.

Why it matters

Orbital data centers are a major driver of SpaceX’s two-trillion-dollar valuation, but the high cost of rockets and satellites means a viable space-based cloud is likely a story for the 2030s.