Amazon Challenges SpaceX's Orbital Data Center Plan at FCC

In a 17-page filing, Amazon argues SpaceX's application is speculative and incomplete, providing technical details for only three "representative" satellites out of a proposed one million. The company claims the plan lacks the basic radio frequency and orbital parameters the FCC requires to assess interference and space safety risks. Amazon's filing calculates that deploying a million-satellite constellation would take "centuries" and that sustaining it would require replacing 200,000 satellites annually, assuming a five-year lifespan. This replacement rate is more than 44 times the entire global satellite launch output in 2025. SpaceX's architectural solution to the immense heat generated by data centers is to distribute the problem. Instead of one massive station, which would require an impossibly large radiator to dissipate heat in a vacuum, it proposes a million smaller satellites, each managing a smaller, more feasible thermal load. A key engineering hurdle is the "obsolescence trap." GPU performance doubles roughly every two years, but a satellite's operational lifespan is typically five to six years. This means orbital hardware could be several generations behind terrestrial data centers, creating a difficult economic and refresh cycle. This conflict is the latest in the rivalry between Amazon's Project Kuiper and SpaceX's Starlink. Starlink currently operates more than 8,000 satellites, while Kuiper has over 100 in orbit with a mandate to launch half of its planned 3,236-satellite constellation by mid-2026. The FCC itself is adapting to this new era, working to streamline licensing for large constellations from a "bespoke process to an applicant-friendly assembly line." This move anticipates a future crowded with non-geostationary systems, but SpaceX's million-satellite proposal tests the limits of this new framework.