Orbital Data Center Funding Surges as Launch Capacity Tightens

A record year, and a new bottleneck

For years, the standard objection to space-based computing was money: building and flying hardware was too expensive and too slow for venture timelines. In 2026 that objection has largely faded. Private funding for space and satellite startups reached $20.3 billion with four months of the year still left. That is the highest annual total Crunchbase has recorded for the sector, and it does not include SpaceX's public offering. 4 Orbital data centers, meaning satellites built to run AI workloads in space, have emerged as a distinct investment category within that surge. 4

The constraint is now shifting. The clearest evidence comes from founders who say their hardest problem is securing a ride to orbit, not raising capital.

The rounds

The largest recent deal went to K2 Space. The Torrance, California company raised a $500 million Series D on July 30 at a $6.8 billion post-money valuation. 1 Kleiner Perkins and ICONIQ Growth co-led, with CapitalG, Lightspeed, Altimeter and ARK Invest participating. 1 It was K2's second nine-figure round in eight months, following a $250 million Series C in December 2025. 1

Investors are backing a cost argument: roughly $15 million per satellite, compared with the hundreds of millions a traditional large-satellite program can cost. 1 K2 flew its first satellite on April 2 and says it has signed more than $1 billion in commercial and government contracts. That backlog is roughly equal to everything it has raised. 1 Tech Times frames K2 as proof that private capital is moving into the orbital compute buildout. 4 Strictly speaking, though, K2 is a satellite manufacturer, not a compute operator. Its relevance to space data centers is mainly as cheap supporting infrastructure.

Starcloud is a purer bet on the category. The startup, which builds satellites for in-orbit AI inference, added a $250 million extension to its $170 million Series A from March, at a $2.3 billion valuation. 2 The money will pay for a larger factory and development of Starcloud-3, its biggest orbital data center spacecraft, which is designed to fly on SpaceX's Starship. 2 Starcloud has also asked the FCC for permission to operate 88,000 spacecraft. 2

At the other end of the scale, Orbital raised a $5 million seed round after graduating from a16z's Speedrun accelerator. 3 Founder Euwyn Poon previously started e-scooter company Spin and sold it to Ford. He has no space background but aims to launch 10,000 space data centers once Starship flies regularly. 3 TechCrunch presented the deal as a sign of how far SpaceX has shifted venture attitudes toward long-horizon, capital-intensive space businesses. 3

Why launch is the new scarcity

Starcloud CEO Philip Johnston was blunt about the constraint. He said one of the company's largest costs is now locking in launch capacity, and that supply is tight because SpaceX's Falcon 9 program is scheduled to end in 2028. 2 "We're going to need to book an enormous amount of launch," he said. 2 He added that Starcloud wants Starship contracts as early as possible. 2

That changes how to read these rounds. Starcloud's extension is partly a war chest for buying rocket slots in a seller's market. Orbital's plan, by its own description, depends on a vehicle that is not yet flying on a routine schedule. 3 Both companies depend on a single provider's next-generation rocket. That concentration risk does not appear in the valuations.

The SpaceX halo

The backdrop is SpaceX's June 12 Nasdaq debut. Shares priced at $135, raising about $75 billion in the largest IPO in U.S. history, at a valuation of roughly $1.77 trillion. 4 Tech Times argues the offering showed institutions that space companies can attract real demand at scale. 4 It also cites a Goldman Sachs projection of a $1.8 trillion space economy by 2035. 4 TechCrunch makes a similar point more narrowly: SpaceX's success has made investors comfortable funding first-time space founders. 3

The two outlets differ somewhat in tone. Tech Times presents a broad, momentum-driven boom. 4 TechCrunch's reporting highlights the practical dependencies underneath it, particularly launch. 23

The takeaway

The money is real, and at least some of it rests on something solid. K2's flight history and contract backlog are the strongest evidence of that. 1 Orbital compute as a business, however, is still mostly a plan to be executed on rockets that one company controls. The binding constraint for this sector has moved from capital to manifest space. The companies most likely to succeed may be the ones that convert today's funding into signed launch contracts first, not the ones with the biggest valuations. Until Starship flies routinely, the field's growth will be limited by launch supply rather than by investor enthusiasm.