FCC reviews Musk's orbital AI data center plan, SpaceX plans to deploy millions of computing satellites

The U.S. Federal Communications Commission (FCC) has officially launched a public review of a groundbreaking proposal by SpaceX. The plan aims to deploy large-scale “Orbital AI Data Centers” in low Earth orbit, sending energy-intensive AI computing tasks into space to support the training of xAI models, including Grok. This concept is seen as a key step for Elon Musk in building a “space-based computing network.”

According to submitted documents, SpaceX plans to operate a new satellite constellation at altitudes of approximately 500 to 2000 kilometers, interconnected via laser optical links, and connected to the existing Starlink network to enable data processing and routing in orbit. The project is extremely large-scale and could potentially expand to hundreds of thousands or even millions of “data center-style” satellites in the future.

This review occurs after Musk integrated xAI into SpaceX, marking a deep integration of rocket launches, satellite communications, and AI computing power. SpaceX stated in the documents that the orbital computing system could provide more energy-efficient AI processing capabilities for global businesses and governments, while also alleviating the massive power and cooling demands of ground-based data centers.

Data shows that in 2024 alone, the electricity consumption of AI data centers in the U.S. will reach approximately 183 TWh, approaching the annual electricity consumption of a medium-sized country. SpaceX believes that relying on nearly continuous solar power and reducing dependence on water cooling systems makes the space-based solution more sustainable in the long term.

The FCC has opened a public comment period, which will run until March 6. During this time, environmental organizations, research institutions, and industry participants can submit their opinions. Although the agency has previously approved Starlink expansion, the orbital data center plan still faces multiple regulatory challenges, including competition, space debris, and spectrum resource management. The outcome of this review could determine the future direction of global AI computing infrastructure.