Google Project Suncatcher: Is the Future of AI Compute in Space?

What if the next generation of AI data centers isn't built on Earth?

That is the idea behind Google Project Suncatcher, one of Google's more ambitious infrastructure experiments. The project explores whether solar-powered satellites carrying Google's Tensor Processing Units, or TPUs, could eventually form a large-scale computing network in orbit.

Google has now announced that Project Suncatcher is preparing for its first orbital test. The company plans to launch prototype satellites to evaluate how its AI hardware performs in space.

The concept sounds futuristic, but there is a practical reason behind it.

Solar panels in certain low-Earth orbits can receive sunlight for much longer periods than solar installations on Earth. Google estimates that solar panels in the right orbit could generate up to eight times more power than comparable panels on Earth.

The bigger vision is a network of interconnected satellites.

Instead of building increasingly massive data centers on land, AI workloads could eventually be distributed across orbital infrastructure. High-speed optical links could allow satellites to communicate and operate as a computing network.

Of course, there are major engineering challenges.

Cooling is different in space. Hardware must survive radiation. Launch costs remain significant. Communication latency and bandwidth also matter.

Google's earlier plan called for prototype satellites to be launched with Planet, with testing expected around 2027. The company has also discussed potential collaboration with SpaceX.

So Project Suncatcher is not an orbital AI data center yet.

It is an experiment into whether space could eventually become another layer of the world's computing infrastructure.

And as AI demand continues to push electricity, land and cooling requirements higher, the question is becoming surprisingly practical:

How far will companies go to find the next source of compute?