Google executed an AI workload on a TPU chip in low Earth orbit, utilizing the Gemma 4 model to generate a space-themed haiku that was beamed back during a Svalbard ground station pass.
This Project Suncatcher milestone tests hardware resilience and execution capabilities in space.
The first communication from space
During a satellite contact with a ground station, the Gemma 4 artificial intelligence model was asked to create a short poem. As the satellite passed over the Svalbard archipelago in Norway's Arctic, the response was transmitted back to Earth.
The satellite, the first prototype of Project Suncatcher, was launched on October 1, in partnership with Planet and the SpaceX launch service. Google said the success proves its specialized chips can perform tasks in space, paving the way for twin connected satellites expected to be operational by 2027.
The technical challenges and the future
Google admits that there are significant obstacles, such as cosmic radiation, solar storms, communication between satellites, and the difficulty of cooling the devices without regular airflow. As the company points out, the first missions are mainly done to see what works well and identify potential weaknesses for future missions.
↳
READ ALSO
HACKS.GR / RELATED
AI Unlocked
OpenAI lowers funding goal to $20 billion
↗
Google's plans include fleets of more than 80 satellites that would fly in close proximity and communicate with each other. The company is also reportedly meeting with designers to potentially build a custom satellite the size of a football field.
Competition in space
The idea of using space for computing is also attracting other companies. In January 2026, SpaceX applied for regulatory approval to launch up to a million satellites to create a massive network in space. In addition, Amazon and Blue Origin founder Jeff Bezos has predicted that huge data centers will be built in space within the next two decades.
>Stay connected with Hacks.gr
Set us as a preferred source on Google and find us on social media.
Preferred source
↗
end of article
Top comments (0)