Google's Project Suncatcher prototype satellite reaches orbit to test AI chips in space

Google has confirmed that its Project Suncatcher prototype satellite is in orbit. In a post dated 1 October 2026, Travis Beals, a senior director at Google, wrote that the team has confirmed contact and that the satellite is operating as expected.[1]

The satellite launched on SpaceX's Transporter-18 rideshare mission and was built in partnership with Planet. Google describes the project as research into whether space could one day host machine learning infrastructure. That is a question, not a product.[1]

Google says the team will gather in-orbit data on how its TPUs handle the physical stress of spaceflight and the radiation and thermal extremes of space. Those are the three hazards that decide whether chips built for air-conditioned halls can survive above the atmosphere.[1]

The company also published a peer-reviewed paper in the journal Joule describing the research behind the mission. It says it will use the orbital data to refine its designs and share more as the mission unfolds. The post does not say how many chips are aboard or how long the mission will run.[1]

In a separate explainer dated 24 September 2026, Google argues that solar panels in orbit can receive up to eight times more power than on the ground. That is Google's claim, and the post does not show the calculation for a working computing satellite.[2]

The same explainer reports ground tests before launch. Components were shaken at forces of 50 to 100 g, and Google says its Trillium TPUs survived radiation exposure equal to a five-year space mission in tests at the University of California, Davis. Heat pipes and radiators were tested in thermal vacuum chambers.[2]

Google says later missions planned for 2027 will test laser links between satellites in a cluster. Those missions are plans, not commitments with dates and budgets.[2]

A working prototype shows that a satellite can fly and talk to its operators. It does not show that thousands of chips can be launched, cooled, networked and repaired at a cost that beats a building on land. Google's post makes no such claim and gives no cost figure.[1][2]

Our reading is that the result to watch is the radiation and heat data, because those numbers decide whether the idea is engineering or only a concept. Google has not said when it will publish them.[1]

Malaysia is a heavy user of the problem Suncatcher tries to solve. MIDA, the national investment agency, said in March 2025 that data centre demand could pass 5,000 MW by 2035, about 40 per cent of Peninsular Malaysia's current power capacity. It also reported RM184.7 billion of data centre investment from 2021 to December 2024.[3]

Google itself is building in Selangor. Its 2024 announcement named a US$2 billion data centre and cloud region at Elmina Business Park, with water cooling it says uses about 10 per cent less energy than air cooling. Orbital computing is a different bet from that building, and neither Google source links the two.[4]

MIDA points to measures on power usage and water usage efficiency, and to schemes that let operators buy renewable power directly. Those are the levers available to Malaysian operators now. A satellite is not one of them.[3]