Chole Ladwig/PitchBook News
“Current advances in AI are dependent on large terrestrial data centers, which require immense amounts of power and cooling,” wrote Elon Musk when he announced that he’s combining SpaceX with xAI. “The only logical solution, therefore, is to transport these resource-intensive efforts to a location with vast power and space.”
For investors whose previous bets on getting data centers in space were dismissed as outrageous, Musk’s rationale provides a shot of legitimacy and validation of the endeavor.
“When we made this investment in 2023, we were laughed off Twitter as this being too audacious an idea,” said Scott Hartley, co-founder and GP at Everywhere Ventures, a pre-seed firm that has backed orbital data center startup Starcloud. “Now you’re having Jensen [Huang], Elon, and Jeff Bezos talk about not just the cost-benefit, but the Earth benefit of putting data centers in space, and so it moves people towards this idea of it being less audacious and more feasible.”
VC space-tech investing has rebounded over the past 12 months, according to PitchBook data. From 2024 to 2025, deal value increased 65% from $6.7 billion to $11.1 billion. And while other categories have witnessed a drop in deal count as venture capital has concentrated into fewer companies, deal count has been on the rise since 2021.
Starcloud and SpaceX aren’t the only startups tackling the challenge. There’s Type One Ventures-backed Axiom Space building a commercial space station and data centers, Lonestar, a startup looking to land data storage systems on the moon, and Sophia Space, an aerospace data center startup with pre-seed financing, to name a few.
And the industry is already showing some promise. In October, Starcloud launched a satellite with an Nvidia H100 GPU onboard. The startup was able to train and query Gemma, an open-source large language model from Google, using the satellite.
But sprinkling data centers around Earth’s orbit is still far from guaranteed. Launching any payload from Earth is still extremely costly and risky. And there are still some unanswered questions about the scientific feasibility of this at the scale needed to support the world’s AI appetite.
“What Elon’s talking about, that’s 2035 at the earliest,” said Brad Harrison, managing partner at Scout Ventures and an investor in Lonestar.
Related read: A leading astrophysicist has some questions for Elon
Houston, no problem
Still, investors are not deterred, seeing orbital data centers as an inevitable answer to the current environmental and economic challenges of providing the infrastructure the AI revolution needs. And in the AI age, speed means everything.
“You don’t have to deal with the EPA [Environmental Protection Agency], you don’t have to deal with all the local NIMBYism,” said Michael Mealling, GP at Starbridge Venture Capital, a space tech-focused firm with investments in SpaceX and Axiom Space. “You have a lot of rural America even saying, ‘we don’t want this stuff here because it drives up costs.’ The need to be fast at scale is where this problem is coming from.”
Outside the (hyper)loop
Investors don’t see SpaceX alone dominating the orbital data center vertical. By integrating X, xAI, and SpaceX into one entity, Musk has created a closed ecosystem akin to what Apple does with its products.
But some customers might want to use a space-based cloud provider compatible with multiple large language models, not just xAI’s Grok.
“People want flexibility, just look at how AI companies are building, coordinating and orchestrating with multiple LLMs,” said Hartley. “You can have access to a bunch of LLMs that Elon doesn’t own. I think that’s the opportunity here.”
The big gamble
Investors are also betting that it’ll be more cost-effective to get data centers in space than build them on Earth—eventually. Advances in rocket technology are reducing the cost of space travel.
And the big bet from investors is that this trend continues, making “data centers in space” a no-brainer.
“AI adoption is outpacing the CapEx ability to build facilities to support the ecosystem,” Hartley said. “There’s this break-even point where the cost of putting things in space becomes worth it. And it is a technical challenge? Absolutely. But there’s this point where we start to move people to this being an economically feasible idea. And I think that movement is happening.”
Additional reporting by Rosie Bradbury.
Learn about our editorial standards.
