Starcloud Closes $250M to Scale Orbital AI

The big picture: Starcloud, a startup building satellites capable of performing AI inference in orbit, secured a $250 million extension to its Series A round, bringing total funding to $420 million and valuing the company at $2.3 billion. Manhattan West Ventures led the round, with participation from Nvidia, Cisco, Benchmark, EQT, Soma, NFX, 776, Cedar Capital, Goanna Capital, and Standard Capital.

Why it matters:

• Orbital AI compute: Starcloud is pioneering AI inference directly in space, operating the first reported Nvidia H100 GPU in orbit and collaborating on Nvidia’s purpose-built Vera Rubin Space-1 chip.
• Scaling infrastructure: The new capital will fund a larger manufacturing facility and the development of Starcloud-3, a major orbital data center spacecraft designed for SpaceX’s Starship.
• Strategic validation: Nvidia’s investment underscores the technical viability and market potential of Starcloud’s approach to distributed AI infrastructure.

How it works:

• In-orbit processing: Starcloud’s satellites host powerful GPUs like the Nvidia H100, enabling AI model inference directly from space-collected data, reducing latency and bandwidth needs.
• Next-gen spacecraft: Development of Starcloud-3 aims to create larger orbital data centers, leveraging advanced launch capabilities like SpaceX Starship for deployment.
• Government applications: Upcoming Starcloud-2 satellites in 2027 will support orbital inference for U.S. government agencies, demonstrating dual-use potential.

The catch: The ambitious plan to deploy 88,000 spacecraft faces significant hurdles, including the tightening market for launch capacity and the long-term cost implications of securing future contracts, especially with SpaceX’s Falcon 9 program slated for retirement.