Two CubeSats orbiting around Earth after being deployed from the ISS. Google is looking at ways satellites can be used for data centres of the future. Credit: NASA
The sun produces more power than 100 trillion times humanity’s entire electricity generation. In orbit, solar panels can be eight times more productive than their Earth-bound counterparts, generating energy almost continuously without the need for heavy battery storage. These facts have led a team of Google researchers to ask what if the best place to scale artificial intelligence isn’t on Earth at all, but in space?
Project Suncatcher, Google’s latest space mission, envisions constellations of solar-powered satellites equipped with processors and connected by laser-based optical links. The concept tackles one of AI’s most pressing challenges, the enormous energy demands of large-scale machine learning systems, by tapping directly into the solar system’s ultimate power source. A new research paper published by Google describes their progress toward addressing the technical challenges.
This artist’s concept shows the Optical Payload for Lasercomm Science (OPALS) and its laser that will beam data to Earth from the International Space Station. This same concept will be used to connect satellites in the new space-based data centers. Credit : NASA/JPL
The proposed system would operate in a sun-synchronous low Earth orbit, where satellites remain in almost constant sunlight. This orbital choice maximizes solar energy collection while minimizing battery requirements. However, making space-based AI infrastructure viable requires solving several formidable engineering challenges.
The first involves achieving data center-scale communication speeds between satellites. Large machine learning workloads require distributing tasks across numerous processors with high bandwidth, low latency connections. Delivering performance which is comparable to Earth-based data centers requires links supporting tens of terabits per second between satellites.
Google’s analysis suggests this should be achievable using dense wavelength division multiplexing and spatial multiplexing technology, but only if satellites fly in extremely tight formation, separated by kilometers or less. The research team has already validated this approach with a bench-scale demonstration that successfully achieved 1.6 terabits per second total transmission.
Flying satellites in such tight formation presents its own challenge. At their planned altitude of around 650 kilometers, satellites positioned less than a kilometer apart would require careful orbital management. Google developed sophisticated physics simulations to analyze how Earth’s non-gravitational field and atmospheric drag would affect these tightly clustered constellations. Their models indicate that only modest station-keeping maneuvers should be needed to maintain stable formations.
Artist impression of the STEREO Observatory spacecraft during solar panel deploy. Satellites in orbit for data center operation will need to be information at low altitude and will therefore need careful orbital control. Credit : NASA/Johns Hopkins University Applied Physics Laboratory
Perhaps surprisingly, Google’s TPU processors appear remarkably resilient to space conditions. Testing of their Trillium v6e Cloud TPU showed the chips could withstand cumulative radiation doses nearly three times higher than expected over a five-year mission before showing irregularities. The High Bandwidth Memory systems proved most sensitive but only began experiencing issues after doses of 2 kilorads, well above the anticipated 750 rads for a shielded five-year mission.
Whether this all makes financial sense depends heavily on launch costs continuing their decline. Google’s analysis suggests that with further improvements in launch technology, costs could fall below $200 per kilogram by the mid-2030s. At that price, launching and operating a space-based data center could become roughly comparable to the energy costs of an equivalent Earth-based facility.
Citation:
Google’s plan for space-based computing (2025, November 12)
retrieved 12 November 2025
from https://phys.org/news/2025-11-google-space-based.html
This document is subject to copyright. Apart from any fair dealing for the purpose of private study or research, no
part may be reproduced without the written permission. The content is provided for information purposes only.
https://phys.org/news/2025-11-google-space-based.html
