Musk Wants a Million Data Center Satellites, Bezos Wants 51,600. Scientists Want to Know Why

SpaceX, Blue Origin, and a growing field of startups are racing to build orbital data centers, with Elon Musk's vision calling for a constellation of roughly one million satellites dedicated to AI computing and Jeff Bezos's Blue Origin targeting a more modest 51,600, The Next Web reported -- a disparity that has researchers raising fundamental questions about the feasibility and environmental impact of space-based compute. The pitch behind orbital data centers rests on a simple premise: AI needs more power than terrestrial grids can supply, and in space the sun never sets, making solar energy a continuous and essentially free resource. Moving compute infrastructure into orbit would, in theory, sidestep the power constraint that is stalling half of planned US data center builds. SpaceX has the launch infrastructure advantage to make such ambitions conceivable. Blue Origin has been developing heavy-lift capability with its New Glenn rocket. A wave of startups including Lumen Orbit and others have raised early funding to explore on-orbit computing. Scientists and astronomers have pushed back on the orbital compute vision on multiple fronts. Astronomers already concerned about Starlink's interference with ground-based telescopes warn that millions of compute satellites would exacerbate light pollution and radio frequency interference severely. Space debris researchers note that the collision risk from large constellations compounds geometrically with satellite count. The practical engineering challenges are also significant. Heat dissipation in vacuum is difficult, orbital compute requires radiation-hardened hardware, and latency from low Earth orbit adds constraints for real-time applications. Whether space-based data centers become a serious infrastructure category or remain an aspirational narrative remains to be seen.