Google and SpaceX Explore Space-Based AI Data Centers Amid Environmental Concerns
Google and SpaceX are exploring space-based computing as a potential way to reduce the electricity and water demands of terrestrial AI data centers. Researchers caution that orbital data centers could create new environmental problems, including heat-management challenges, space debris, light pollution and launch-related emissions.
Google is testing AI computing in orbit through Project Suncatcher, while SpaceX is developing plans for satellites that could eventually support large-scale space-based computing. The companies say orbital systems could benefit from abundant solar power and reduce the need for water-intensive cooling and new power plants on Earth.
Duke University researcher Arun Kannawadi said the approach would still require sophisticated heat dissipation and could produce significant environmental risks. A constellation approaching 1 million satellites could interfere with astronomical observations, brighten the night sky, increase collision risks and generate orbital debris, while manufacturing and launching the satellites would create additional emissions and resource demands.
The technology remains experimental and does not address near-term electricity demand from North Carolina's terrestrial data center industry. Duke Energy projects data centers could represent about 10% of its electricity sales across the Carolinas by 2030 and is planning billions of dollars in new generation and infrastructure, prompting concerns about customer costs and fossil-fuel emissions.