Google has launched Project Suncatcher, an initiative to explore the feasibility of deploying artificial intelligence (AI) capabilities in space. The plan involves sending swarms of low-Earth orbit satellites equipped with Google’s AI accelerator chips, known as Tensor Processing Units (TPUs), to handle tasks such as training AI models, content generation, synthetic speech, computer vision, and predictive analytics.
Google describes Project Suncatcher as a “moonshot” aiming to equip solar-powered satellite constellations with TPUs and free-space optical links to expand machine learning compute beyond Earth. The company acknowledges significant engineering challenges, including managing thermal conditions and ensuring system reliability in orbit. CEO Sundar Pichai confirmed initial tests show TPUs can withstand harsh space radiation.
Demand for AI compute is rapidly increasing, driving growth in data centers that consume massive amounts of electricity and water, raising environmental concerns. Travis Beals, senior director of Google’s Paradigms of Intelligence research team, said the project seeks a scalable solution to meet this growing demand without the limitations of terrestrial infrastructure.
According to projections cited by MIT Technology Review, AI energy consumption could reach that of 22 percent of US households by 2028. Project Suncatcher aims to address these sustainability challenges by moving AI workloads to orbit.