SpaceX wants to fill Earth orbit with a million datacenter satellites

Steven

SpaceX wants to fill Earth orbit with a million datacenter satellites

Summary

SpaceX has filed with the FCC to launch up to one million “orbital datacentre” satellites, and the FCC’s Space Bureau has accepted the application for public comment. The proposed constellation would operate in multiple shells between about 500 km and 2,000 km altitude, use high-bandwidth optical links, and route traffic via Starlink to authorised ground stations. SpaceX claims very high compute density (around 100 kW per metric ton of compute) and plans to rely on Starship for mass launches.

Key Points

  • SpaceX applied to the FCC on 30 January for permission to deploy up to 1,000,000 datacentre satellites; the FCC opened the application for public comment.
  • The proposed satellites would occupy multiple orbital shells between ~500 km and ~2,000 km altitude and use optical inter-satellite links plus Starlink for ground traffic.
  • SpaceX claims roughly 100 kW of power per metric ton for onboard computing; many technical details are sparse and speculative.
  • The plan depends heavily on Starship for bulk launches — a vehicle that has had limited operational success to date.
  • A constellation of this scale would dramatically increase orbital crowding, raise Kessler Syndrome risks, and further interfere with astronomical observations.
  • Experts stress the need for active debris removal or “tow-truck” satellites to mitigate failed units; existing deorbit practices may be insufficient at scale.
  • The FCC is soliciting public comment, with a deadline noted in the filing (deadline for comments: 6 March 2026).

Content summary

The Register reports that SpaceX’s filing frames the massive constellation as an “orbital datacentre” project intended to scale compute in space. The filing references optical links and integration with Starlink, and invokes grand ambitions (even citing Kardashev civilisation terminology). Harvard astrophysicist Jonathan McDowell and others warn the proposal would massively increase the number of active satellites in orbit — from ~14,500 to potentially over a million — exacerbating collision risks, debris generation and interference with Earth- and space-based astronomy.

Concrete technical detail in SpaceX’s narrative is limited; claims about power-per-tonne and launch/operational efficiencies rest on technologies and launch rates that are not currently realised. The article notes ongoing concerns about current deorbiting rates and highlights industry interest in active removal solutions.

Context and relevance

This proposal, if taken seriously, would be one of the largest planned increases in orbital population in history. It sits at the intersection of cloud/datacentre growth, commercial spaceflight ambition, and long-standing concerns about space sustainability. For astronomers, regulators and satellite operators, the plan raises immediate questions about space traffic management, debris mitigation, liability and the long-term usability of low- and medium-Earth orbit.

Governments and agencies have been wrestling with constellation regulation, collision avoidance and the environmental impact of space activities; an application of this magnitude would force those debates to accelerate and could prompt stricter licensing conditions or technological safeguards if it proceeds.

Why should I read this?

This is bonkers — and important. If you care about the night sky, satellite safety, regulation or the future shape of cloud compute, this one filing could change the game. The article saves you the time of digging through the FCC paperwork: it flags risks (debris, astronomy interference), big assumptions (Starship scale, power claims), and a fast-approaching public comment window so you can say something if you want.

Source

Source: https://www.theregister.com/2026/02/05/spacex_1m_satellite_datacenter/