Experts question near-term viability of orbital data centers as SpaceX’s Starship faces delays

Sam Altman and Elon Musk engaged in a public exchange over the weekend, each challenging the other’s credibility regarding the feasibility of orbital data centers for AI inference tasks. Altman characterized Musk’s promotion of short-term space data centers as aimed at public market investors, while Musk accused Altman of being a scammer.

Experts consulted by TechCrunch, including entrepreneurs in the space data center sector and engineers at Google’s orbital compute project, uniformly expressed skepticism about the near-term viability of the concept. They argued that orbital AI inference centers will not become a significant business until launch costs fall dramatically and high-powered satellites can be produced at scale and low cost.

SpaceX’s Starship, the rocket central to Musk’s space data center ambitions, remains years away from achieving operational reusability, according to expert assessments and company disclosures. SpaceX’s IPO road show materials explicitly stated that Starship may not be fully reusable in the near term, requiring the discard of second stages during launches, which would undermine the economic case for space data centers.

Even if Starship achieves its 13th test flight as soon as July 16, 2026, analysts note that routine, cost-effective reuse is likely years away. Meanwhile, SpaceX’s priorities—such as NASA commitments and expanding the Starlink network—are expected to take precedence over launching space data centers at scale.

Musk’s response that SpaceX could begin flying space data centers next year was met with skepticism, as the core challenge remains not launch capability but the ability to manufacture and operate satellites at scale and low cost—an outcome experts suggest is unlikely before the 2030s.