Starcloud, a Y Combinator graduate, has secured $170 million in Series A funding, valuing the space compute company at $1.1 billion and establishing it as a unicorn. Led by Benchmark and EQT Ventures, the round closes just 17 months after the company's demo day, bringing its total capital raised to $200 million. This investment underscores growing interest in relocating data centers to orbit, driven by terrestrial resource constraints and political hurdles, though the sector relies on unproven technology and massive capital expenditure. Since launching its first satellite equipped with an Nvidia H100 GPU in November 2025, Starcloud is preparing to deploy Starcloud 2 later this year. This next iteration will feature multiple GPUs, including an Nvidia Blackwell chip and an AWS server blade, alongside specialized bitcoin mining hardware. The company's long-term vision involves Starcloud 3, a 200-kilowatt, three-ton data center spacecraft designed to launch from SpaceX's Starship using a specialized deployment system. CEO Philip Johnston projects that Starcloud 3 will eventually achieve cost competitiveness with Earth-based facilities, targeting power costs of approximately $0.05 per kilowatt-hour. However, this milestone depends on commercial Starship flights reaching a high operational cadence, a reality Johnston expects by 2028 or 2029. Until then, the company may rely on SpaceX's Falcon 9 rocket for smaller missions, accepting higher energy costs in the interim. The company operates on a dual business model: initially selling processing power to other orbital assets, such as analyzing radar data for Capella Space, and later serving as a distributed infrastructure provider for terrestrial workloads. This nascent industry faces significant technical hurdles, including efficient power generation, cooling for high-performance chips, and synchronization across multiple spacecraft. Current space data centers handle relatively small workloads compared to the gigawatts required for large-scale AI training on Earth. Starcloud claims a lead in this field, having successfully trained an AI model and run Google's Gemini software in orbit. While their initial H100 chip proved the viability of terrestrial hardware in space, subsequent testing revealed failures, such as an Nvidia A6000 malfunctioning during launch, providing valuable data for future designs. Starcloud 2 will incorporate the largest deployable radiator ever flown on a private satellite to manage heat. Competition is intensifying, with Aetherflux, Google's Project Suncatcher, and Aethero also developing space-based compute solutions. SpaceX itself has sought approval to build a million-satellite network for distributed computing, potentially serving its own Grok and Tesla workloads. Johnston distinguishes Starcloud's approach as focused on energy and infrastructure for third-party services rather than proprietary cloud computing, suggesting a path for coexistence rather than direct conflict with the space giant. Despite the optimism, the sector remains in its infancy, with the number of advanced GPUs in orbit numbering in the dozens against millions deployed on Earth.