Sam Altman and Elon Musk exchanged barbed posts on social media over the weekend, drawing new attention to the gap between vision and reality in the space computing business.
After Musk accused him of being a fraud, Altman said, “Homeboy, you’re selling public market investors on short-term space data centers.”
“Homeboy” aside, Altman is saying what many experts have concluded but public market investors seem to be ignoring: Space data centers won’t become a serious business anytime soon.
SpaceX’s plans to launch a fleet of orbiting data centers to perform AI inference tasks are a major driver of the company’s $2 trillion valuation. Bullish analysts say the potential for its processing power to fuel SpaceXAI’s models or serve as an orbital neocloud is unprecedented in the AI boom.
But if you talk to experts in the field, whether they’re entrepreneurs who’ve started other space data center startups, a team at Google developing the company’s orbital computing project, or an engineer who crunched numbers as a hobby, you’ll likely find the same answer. This won’t be a big problem until we have much cheaper rockets and the ability to mass-produce low-cost, high-power satellites.
Musk’s answer to this is easy to predict. Starship, SpaceX’s massive new rocket, is scheduled to make its 13th test flight as early as July 16. If Musk’s team can get the rocket to a state where it can fly multiple times, the data center business case could be solved.
But even if the company succeeds in recovering both rocket stages on this test flight, reusable flight is likely still years away, and the launch of a space data center will likely take a backseat to SpaceX’s commitment to NASA and the creation of its own Starlink network.
SpaceX also acknowledged at its IPO roadshow that Starship may not be fully reusable in the short term, requiring each launch to throw a second stage, which would be a huge blow to economic space data centers.
That’s why Musk’s response that “we’ll start flying next year” is a little off base. There’s no question that SpaceX could launch a satellite with high-speed data processing next year, but the big question is when it will be able to launch and manufacture it at scale. And that’s probably going to be a problem for the 2030s.
If you buy through links in our articles, we may earn a small commission. This does not affect editorial independence.
