Experts Already Agree with Sam Altman’s Space Data Center Critique
▼ Summary
– Sam Altman and Elon Musk exchanged public criticism over space data centers, with Altman accusing Musk of selling a short-term vision to public investors.
– SpaceX’s plans for orbital data centers for AI inference are a key driver of its $2 trillion valuation, but experts widely agree the business is not viable soon.
– The feasibility of space data centers depends on much cheaper rockets and mass production of low-cost, high-powered satellites, which are not yet available.
– Starship’s potential reusability could enable the business case, but operational reuse is likely years away, and SpaceX’s NASA and Starlink commitments take priority.
– SpaceX may launch a prototype satellite next year, but scaling to economical, mass-produced space data centers is not expected until the 2030s.
Sam Altman and Elon Musk exchanged pointed remarks on social media over the weekend, shining a fresh spotlight on the growing disconnect between the hype and the hard realities of the space data center industry.
When Musk accused Altman of being a scammer, Altman fired back, writing, “homeboy you’re the one selling public market investors on short-term space datacenters.” The jab aside, Altman’s underlying point aligns with what numerous subject-matter experts have quietly concluded but public market investors appear to be overlooking: orbital data centers are not poised to become a serious commercial venture anytime soon.
SpaceX’s ambitious plans to deploy a constellation of orbital data centers for AI inference tasks have fueled much of the company’s staggering $2 trillion valuation. Optimistic analysts argue that the potential for this processing power to support SpaceXAI’s own models or function as an orbital neocloud could be unprecedented in the AI boom. Yet, conversations with entrepreneurs behind other space data center startups, engineers at Google working on similar orbital compute projects, and independent analysts who have crunched the numbers all yield the same verdict: this concept will not make a meaningful impact until we achieve far cheaper rockets and the ability to mass-produce high-powered satellites at low cost.
Musk’s predictable counterargument rests on Starship, SpaceX’s colossal new rocket, which is slated for its 13th test flight as soon as July 16. If the team can demonstrate rapid reusability, the business case for space data centers could theoretically close. However, even a successful recovery of both rocket stages on this flight does not guarantee operational reusable flights anytime soon. Space data center launches are likely to take a backseat to SpaceX’s existing commitments to NASA and the ongoing expansion of its Starlink network.
SpaceX itself acknowledged during its IPO road show that Starship may not achieve full reusability in the near term. The company may need to discard each second stage after every launch, a reality that would effectively kill the economics of orbital data centers. That is why Musk’s retort,“We start flying them next year”,feels underwhelming. While SpaceX could certainly launch a satellite capable of high-speed data processing within the next year, the critical question is when it can launch and manufacture these systems at scale. The honest answer likely points to the 2030s.
(Source: TechCrunch)
