Project Suncatcher is a long-term research effort to explore running AI computing in space. After launching its prototype satellite on October 1, the team completed initial commissioning, powered up the onboard host and TPU trays, and ran early machine-learning tests. The key milestone came when a ground station sent a prompt to the satellite’s Gemma 4 open model and received its first AI-generated response from an orbiting, data-center-class TPU. During a pass over Svalbard, Norway, the model produced a haiku about space, confirming that the system could accept a prompt, compute an answer in orbit, and send it back to Earth.
The team then tested other prompts, including requests for space-themed limericks and reports on the payload’s technical status. These demonstrations establish that the hardware can perform basic AI workloads in orbit, but the next challenge is to determine whether it can do so reliably in harsh conditions. Planned tests will examine TPU performance under thermal stress and radiation exposure. If the chips prove dependable, the project can advance toward its next milestone: two interconnected satellites, targeted for 2027. The successful response is an initial technical step toward the broader goal of exploring AI compute infrastructure in space, not a completed operational system.
Summary generated by AI from the linked article. hn.today is not affiliated with Hacker News or Y Combinator.