Google has launched the first prototype satellite for Project Suncatcher, a long-term research initiative exploring whether space could eventually support scalable machine learning infrastructure.
The prototype launched into orbit on October 1 aboard SpaceX’s Transporter-18 rideshare mission. Google developed the satellite in partnership with Planet.
Google confirmed that its team successfully established contact with the satellite after launch and that the spacecraft is operating as expected.
The mission represents the first in-orbit test for Project Suncatcher, which is investigating whether space-based computing infrastructure could eventually support demanding machine learning workloads.
A primary objective of the prototype mission is to test Google’s Tensor Processing Units under actual spaceflight conditions.
Over the coming weeks, Google’s research team plans to collect data on how the TPUs respond to the physical stresses associated with launch and operation in orbit.
Researchers will also evaluate how the hardware performs when exposed to radiation and the extreme thermal conditions encountered in space.
The tests are intended to provide real-world information that cannot be fully replicated on Earth and help Google refine the hardware and system designs being considered for future Project Suncatcher development.
Google has also published a peer-reviewed paper in the scientific journal Joule detailing the research underlying the project and the prototype mission.
Project Suncatcher is being developed as a research moonshot rather than a near-term commercial product. The project is examining the technical requirements and limitations involved in moving portions of AI and machine learning infrastructure beyond terrestrial data centers.
Google said the satellite experiments will provide information that can be incorporated into future designs as the company continues evaluating the feasibility of scalable machine learning infrastructure in space.
