orbital data center
In plain wordsA proposed computer center circling the Earth, powered by almost constant sunlight, that would run AI without using land, water or power grids on the ground.
A data center placed on satellites in Earth orbit, proposed as a way to meet the growing electricity demand of AI computing with near-continuous solar power while sidestepping terrestrial limits on land, grid connections and water. Designs envisage constellations of solar-powered satellites carrying AI accelerators and linked by free-space laser connections. Obstacles include radiation damage to chips, shedding heat in a vacuum where it can only be radiated away, launch costs, maintenance and latency. In November 2025 Google reported that its Trillium TPUsTPU, abbr.: Google's own family of chips built specifically for AI, used to train and run its Gemini models and rented to other companies through its cloud. had withstood radiation doses well above those expected over a five-year mission, and planned to launch two prototype satellites with the operator Planet by early 2027; the start-up Starcloud had by then flown a data-center GPU in orbit.
The pitch deck promised an orbital data center by 2030, with cooling still marked 'TBD'.
Origin
Long a speculative idea, it became a serious industry topic in late 2025, when Google announced its Project Suncatcher research program (November 2025) and Starcloud launched a satellite carrying an Nvidia H100 GPU; executives including Elon Musk and Jeff Bezos promoted the concept.
Related
- AI data centerAI data center, n.: A building or campus packed with AI chips, power equipment and cooling, built to train and run AI models. The biggest use as much electricity as a city.
- environmental impact of AIenvironmental impact of AI, n.: The effects of AI on the planet: the electricity, water, materials and land used to make chips, build data centers and train and run models, and the pollution that results.
- TPUTPU, abbr.: Google's own family of chips built specifically for AI, used to train and run its Gemini models and rented to other companies through its cloud.
- computecompute, n.: The raw computing power used to train and run AI, usually measured by how many specialized chips are used and for how long. With data and algorithms, it is a main driver of AI progress.
- AI factoryAI factory, n.: A data center built to turn electricity and data into AI models and AI answers at industrial scale; in Europe, also a public supercomputing hub that helps companies and researchers build AI.
- liquid coolingliquid cooling, n.: Cooling computer chips by running liquid past them instead of blowing air. AI chips now run so hot and are packed so tightly that air alone can no longer keep them cool.
Sources
- Clark, J. (2026). Import AI 474: Platonic mindspace; TPUs in space; Zhipu starts an outer RSI loop. 28 September 2026. jack-clark.net/2026/09/28/import-ai-474-platonic-mindspace-tpus-in-space-zhipu-starts-an-outer-rsi-loop