Data centers are the backbone of the digital economy, powering everything from telehealth visits and online classrooms to the AI tools that help Main Street businesses compete. As Americans rely on these services more and more, our innovators are rethinking where the necessary infrastructure for computing power can live. This week, Google gave us a glimpse of one of the boldest answers yet: a prototype satellite built to test whether AI chips can run in orbit.
What Is Project Suncatcher?
Project Suncatcher is a long-term research moonshot that Google first announced last year. It asks a simple but audacious question: could space one day host data centers? The potential starts with energy. Satellites in low Earth orbit enjoy near-constant sunlight and can generate up to eight times more solar power than panels on the ground, according to Google.
The project’s first real-world test is now heading to orbit. A prototype satellite carrying Google’s Tensor Processing Units (TPUs) — the chips that power its AI models — will fly aboard SpaceX’s Transporter-18 rideshare mission, scheduled for October 1. Google developed the mission in partnership with American satellite company Planet, and it is designed to collect in-orbit data on how the chips handle launch stress, radiation, and extreme temperatures.
This is only step one. Google plans a second milestone in 2027, when it will put two satellites into orbit to test how they communicate, and the team is documenting its progress in a public video series.
Why Look to the Sky?
America’s appetite for AI is growing fast, and so is the infrastructure needed behind it. Gartner projects that global data center electricity use will reach 565 terawatt-hours in 2026, a 26% jump over 2025, with AI-optimized servers accounting for nearly a third of that total. Goldman Sachs Research expects U.S. data center power demand to more than double, from 31 gigawatts in 2025 to 66 gigawatts in 2027.
Demand this strong is a sign that Americans value what these tools deliver — and it is exactly the kind of signal that pushes innovators to find new sources of power, new cooling methods, and entirely new places to build. Project Suncatcher is a vivid example.
Solving the Hard Problems, One Test at a Time
Google is candid that putting AI compute in space will take methodical engineering, not a single leap. The team has broken the challenge into concrete problems, including surviving launch, withstanding radiation, and cooling, and early results are encouraging.
The roughly 10-minute ride to low Earth orbit subjects a spacecraft to intense vibration and sustained forces up to 10 times that of gravity, while individual chips can experience even greater forces up to 50 to 100 g. Google shook its satellite along all three axes to simulate launch conditions, and the hardware held up.
Cosmic rays and solar events can scramble electronics. Google tested its TPUs in a proton beam at UC Davis’ Crocker Nuclear Laboratory while they ran live AI workloads. Initial results demonstrate that its Trillium TPUs withstood a total radiation dose greater than what they would absorb over a five-year space mission.
AI chips also pack a lot of heat into a small space, and in orbit there is no air to carry it away, risking overheating. The team is testing several ways to address this issue, including combining heat pipes and radiators, and has already tested its system in a thermal vacuum chamber that mimics conditions in space.
What This Means for the Future of Compute
Project Suncatcher brings together years of privately funded research, a U.S. chip designer, a U.S. satellite company, and a U.S. launch provider — all with no government mandate required. Google itself likens the effort to early work on autonomous driving and quantum computing, fields that needed years of experimentation before they produced practical systems.
Moratoriums on data center construction, slow-walked permits for new power generation, and a patchwork of state laws governing AI all raise the cost of building the terrestrial infrastructure that funds and informs moonshots like this one. Innovation-killing red tape on the ground makes it harder to reach new milestones in the sky.
The stakes are global. Gartner’s analysts describe power security as the new battleground in the global AI race. The country that answers rising demand with more innovation, not restriction, will lead the next era.