Google Takes Space-Based Computing to New Heights with Project Suncatcher

Google is taking its research into space-based computing to new heights with Project Suncatcher. The tech giant plans to launch a prototype of the system aboard SpaceX's Transporter-18 mission, which will test the hardware in space. This flight marks an important step in Google's effort to explore whether solar-powered satellites can support machine-learning computing.

The project's broader concept involves a system of solar-powered satellites connected by optical links, with the goal of creating a more efficient and scalable way to process data. In laboratory testing, Google Research has already demonstrated the capabilities of its Trillium chip, which is designed for high-bandwidth memory (HBM) applications. The chip was exposed to a 67 MeV proton beam and showed no hard failure attributable to total ionizing dose (TID) through 15 krad(Si).

However, there are still significant engineering challenges to overcome before Project Suncatcher can become a reality. Google Research has identified thermal management, communication with the ground, and reliability in orbit as key areas of focus. The next planned step is a follow-up learning mission with Planet, which aims to launch two prototype satellites by early 2027.