Google Unveils Project Suncatcher to Build AI Data Centers in Space

Google has announced Project Suncatcher, a research initiative to design space-based AI infrastructure using solar-powered satellite constellations equipped with Tensor Processing Units (TPUs), detailed in a company blog post. The company plans to launch two prototype satellites in partnership with Planet by early 2027 to test key technologies for scalable machine learning compute in orbit.

According to the research, the proposed system would operate in a sun-synchronous orbit, allowing near-continuous solar power generation. The satellites would use free-space optical links to transmit data at high bandwidths, forming a distributed compute network. Tests of Google’s Trillium-generation TPUs showed strong radiation tolerance in simulated space conditions.

The company’s analysis suggests that as launch costs continue to fall, space-based data centers could become economically comparable to terrestrial ones by the mid-2030s. The modular satellite design, with units flying just hundreds of meters apart, aims to achieve data center–scale inter-satellite communication performance.

Google identified several challenges still to be addressed, including thermal management, ground communications, and long-term system reliability. The upcoming 2027 mission will serve as a learning experiment to validate these models and hardware in orbit.