Starcloud
Develops orbital data centers that provide GPU compute and AI inference services in space using solar power and passive radiative cooling
Starcloud designs and operates satellites that function as orbital data centers, providing GPU compute infrastructure for artificial intelligence workloads. The company leverages continuous solar energy and the vacuum of space for passive radiative cooling, enabling sustained operation of data center-grade GPUs without traditional cooling systems. Formerly known as Lumen Orbit, the company graduated from Y Combinator's S24 batch.
Starcloud-1, a 60-kilogram demonstration satellite launched in November 2025, carried an NVIDIA H100 GPU into orbit — the first commercial AI-grade processor operated in space — and became the first spacecraft to train an LLM and run Google's Gemma model. The first commercial mission, Starcloud-2, targets a 2027 launch with a GPU cluster, persistent storage, and proprietary thermal and power systems in sun-synchronous orbit. In March 2026, the company filed with the FCC for an 88,000-satellite constellation at 600–850 km altitude with optical inter-satellite links.
Starcloud operates on an infrastructure-as-a-service model through partnerships including Crusoe Cloud and Capella Space. Founded in 2024 and headquartered in Redmond, Washington, the company is backed by NFX, Y Combinator, In-Q-Tel, FUSE, 468 Capital, Andreessen Horowitz, and Sequoia.
Operates a satellite-as-a-service platform that integrates, deploys, and operates customer payloads and AI applications on standardized LEO spacecraft
Operates a constellation of hyperspectral imaging satellites and delivers spectral intelligence for energy, mining, agriculture, and defense monitoring