Starcloud raises $250M as launches tighten

Starcloud raised a $250 million Series A extension, taking its post-money valuation to $2.3 billion, on top of the $170 million Series A it closed in March. Manhattan West Ventures led the round, with Nvidia putting in $25 million alongside Cisco, Benchmark, EQT, and several other funds. The company builds satellites carrying AI inference hardware, it already runs Nvidia H100 GPUs in orbit and is developing purpose-built chips for space deployment.

CEO Philip Johnston tied the raise directly to a launch-capacity squeeze: SpaceX is scheduled to phase out the Falcon 9 by 2028 in favor of the still-unproven Starship, while Blue Origin’s New Glenn and ULA’s Vulcan aren’t flying on a regular cadence and Rocket Lab’s Neutron isn’t available yet. “Launch is pretty constrained right now,” Johnston said, and the fresh capital is meant to lock in launch contracts before the bottleneck gets worse.

What it means for you

Orbital data centers solve a real constraint, terrestrial power and cooling for AI training are both getting harder to site, but this raise is really a bet on securing scarce rocket capacity, not evidence the economics of space-based inference already work at scale. Nvidia’s participation is the more interesting signal than the headline number: a chipmaker putting money into a customer that’s testing its hardware in orbit is a way to keep a toehold in a speculative but plausible future compute market, similar to the logic behind xAI’s Memphis turbine buildout for terrestrial power instead, or Lambda’s GPU leaseback financing, another sign that compute capacity itself is becoming the asset investors want exposure to, wherever it sits. Worth tracking if you’re forecasting compute supply years out, not something that changes anything for a working operator today.