Starcloud Secures $250 Million to Build Orbital AI Data Centers

Manhattan West led the round, with participation from Nvidia and Cisco Investments. The financing puts Starcloud at a $2.3 billion post-money valuation and takes total capital raised since its 2024 founding to $450 million.

The headline is not the round size. It is the bet behind it: investors are funding computing infrastructure before the launch economics are proven.

The capital stack is ahead of the product

Dealroom reports that Nvidia contributed about $25 million, citing a person familiar with the deal. The new financing will support launch procurement, manufacturing capacity, engineering work with Nvidia, and the development of orbital AI data centers.

The current operating claims are narrow but measurable:

• Starcloud-1 launched in November 2025 with an Nvidia H100 GPU.
• The company says it trained the first AI model in space.
• It also says the system ran a version of Google’s Gemini.
• Starcloud is working with Nvidia on the Space-1 Vera Rubin Module, a chip designed for space and scheduled to fly in late 2028.

That is enough to establish a technical demonstration. It is not enough to establish a viable infrastructure business.

The funding is large even by venture standards. Dealroom places the round in the top 1% of Series A deals by size in its sector. That is the relevant comparison. Starcloud is not being financed like a conventional early-stage software company. It is being financed like a strategic infrastructure option.

Scale depends on two unproven numbers

Starcloud has requested permission from the US Federal Communications Commission to operate 88,000 spacecraft. Its target is 20 gigawatts of orbital capacity.

Those figures describe ambition, not deployed capacity. The company’s immediate manufacturing plan is more concrete. Next week, it is expected to move into a 100,000-square-foot facility in Woodinville, Washington, for production of its next-generation Starcloud-3 spacecraft. Starcloud says the site could support output of 100 satellites per week.

The constraint is launch access.

Dealroom reports that SpaceX plans to retire Falcon 9 by 2028, while Starship remains unproven. Blue Origin’s New Glenn and ULA’s Vulcan are also not flying regularly, according to the report. Starcloud expects to book a large amount of launch capacity, and much of the new capital is allocated to that problem.

This creates a hard dependency:

• Manufacturing capacity can be expanded on a fixed site.
• Orbital capacity cannot expand without available launches.
• The business case therefore depends on launch frequency, price, and reliability that Starcloud does not control.

The company is effectively buying exposure to a future launch market while building the rest of the system today.

What builders and investors should track

For operators evaluating the category, the useful metrics are not spacecraft counts or fundraising totals. Track:

• Capital raised: $450 million since 2024.
• Current round: $250 million.
• Post-money valuation: $2.3 billion.
• Target orbital capacity: 20 gigawatts.
• Requested fleet size: 88,000 spacecraft.
• Planned manufacturing rate: 100 satellites per week.
• Key technical milestone: Space-1 Vera Rubin Module, targeted for late 2028.

The gap between those numbers is the story. A $2.3 billion valuation is being assigned to a company whose expansion plan still relies on launch infrastructure with limited regular availability.

Nvidia and Cisco Investments provide strategic validation. They do not remove launch costs, orbital operations risk, or the gap between one demonstrated H100 mission and a 20-gigawatt network.

Verdict: viable as a funded infrastructure option, unproven as an investable operating model. The next decisive evidence will be repeatable launches and production at scale, not another financing round.