The AI revolution is built on a hidden cost: an insatiable thirst for water. Terrestrial datacenters, the backbone of the $3 trillion AI boom, consume millions of gallons of water for cooling. Google’s “Project Suncatcher” is a radical plan to solve this resource crisis by making it obsolete.
Google’s plan, which involves placing AI datacenters in orbit, “minimises impact on terrestrial resources.” By moving to the vacuum of space 400 miles up, the need for water-based cooling is eliminated entirely. This would delink the growth of AI from the depletion of one of Earth’s most critical resources.
This is a key driver for the “moonshot” project. The company’s research highlights the land and water impact of earthbound datacenters as a primary motivation. Space, in this context, is seen as an “unlimited” resource, powered by solar panels that are 8-times more effective.
The challenge, however, is simply shifted. Instead of water cooling, Google’s engineers must now solve for “thermal management” in a vacuum, where heat must be radiated away. This is one of the “significant engineering challenges” the company admits it must overcome.
While the plan faces its own environmental hurdles, such as rocket CO2 emissions and astronomer objections, its core promise is tantalizing. It offers a future where AI can scale without putting further strain on our planet’s stressed land and water supplies.