Google will soon be one step closer to getting AI data centers up and running in outer space.
The tech giant announced Thursday that its Project Suncatcher will launch a prototype satellite into low Earth orbit next week to test how well Google Tensor Processing Units (TPUs) hold up under the rough conditions of space.
“This first launch is about seeing what works, identifying points of failure, and applying those findings to future missions,” the company said in a blog post.
Project Suncatcher is Google’s long-term research moonshot, first announced last year, exploring whether massive AI computing infrastructure could eventually operate in space.
The basic idea is that satellites packed with AI chips could one day take advantage of near-constant sunlight for power and communicate with each other using lasers to handle large AI workloads. Google says satellites in low Earth orbit could generate up to eight times more solar power than on Earth.
This could help ease some of the energy and environmental issues facing terrestrial data centers, which are seeing growing backlash on Earth.
Google is far from alone in looking to space as a solution. Companies like SpaceX and Starcloud are also exploring orbital data centers. Starcloud launched an Nvidia H100 GPU into orbit in November 2025. SpaceX, meanwhile, made space-based AI infrastructure a major part of its IPO pitch earlier this year.
As for Google, its prototype satellite will hitch a ride on SpaceX’s Transporter-18 rideshare mission next week. Developed with aerospace company Planet Labs, the mission will send four Google TPUs into orbit for the first time.
The mission will collect data on how the chips handle the trip to space and low Earth orbit.
For example, Google says the roughly 10-minute rocket ride into low Earth orbit subjects the spacecraft to intense vibrations and forces up to 10 times stronger than Earth’s gravity. Individual components, including the TPU chips, can also briefly experience forces 50 to 100 times stronger than gravity.
Once in orbit, the chips will also face high-levels of radiation from solar events and cosmic rays, which can damage electronics or cause data errors known as bit flips.
Google previously tested its Trillium TPUs at UC Davis’s Crocker Nuclear Laboratory by exposing them to a proton beam while the chips ran AI workloads. The team monitored how errors such as bit flips affected those workloads. According to Google, the chips held up “remarkably well,” surviving a total radiation dose greater than what they would typically encounter during a five-year space mission.
Additionally, AI chips generate a lot of heat, but there is no airflow in the vacuum of space. That means Google can’t rely on the kinds of cooling systems commonly used inside terrestrial data centers.
“We’re working on a number of different approaches for this, including a combination of heat pipes and radiators to cool the chips,” the company said in its blog post. “So far, our team has tested the technology in a thermal vacuum chamber that simulates both the thermal and vacuum environment in space. We’ll see how our new TPU cooling system works in space and refine our designs as we learn more.”
The data gathered from these tests will help shape Project Suncatcher’s future missions. Google already plans to put two satellites into orbit in 2027 and test the high-bandwidth lasers it hopes could eventually connect clusters of satellites.
Read the full article here
