No land, no cooling, no grid: the economics of orbital data centers
Blue Origin's Project Sunrise FCC filing is less about a data center and more about securing orbital slots before someone else does. Blue Origin has filed with the FCC for Project Sunrise, a constellation of up to 51,600 satellites in sun-synchronous orbits between 500 and 1,800 kilometers altitu...

image from FLUX 2.0 Pro
Blue Origin has filed with the FCC for Project Sunrise, a constellation of up to 51,600 satellites in sun-synchronous orbits between 500 and 1,800 kilometers altitude, designed to provide in-space computing services. As Jeff Foust reported for SpaceNews, the company argues that orbital data centers powered by continuous solar energy can deliver compute capacity at lower marginal cost than terrestrial alternatives.
The argument Blue Origin makes in the filing is not wrong in principle. A satellite in sun-synchronous orbit above 500 kilometers has essentially uninterrupted solar power for most of the year, no land acquisition costs, no cooling infrastructure, and no grid constraints. For AI workloads that demand continuous power and can tolerate the latency of space-to-ground links, the physics does favor orbit. The company states that "lack of land or displacement costs, and nonexistent grid infrastructure disparities fundamentally lower the marginal cost of compute capacity compared to terrestrial alternatives."
But the filing is as much about spectrum as it is about satellites. Blue Origin is asking the FCC for a waiver from milestone rules that would require half the constellation in orbit within six years of approval and the full constellation within nine years. Its argument: because it is seeking Ka-band spectrum on a non-interference basis rather than exclusive use, the milestones are unnecessary and it is not "warehousing" spectrum. That argument has a structural implication: the company is not committing to a deployment timeline because it does not yet have one.
Project Sunrise would integrate with TeraWave, Blue Origin's broadband constellation announced in January, primarily through optical intersatellite links. New Glenn is cited as the launch vehicle enabling the system, though no deployment schedule is specified. The company says it will deorbit satellites within five years of end of life and work with the astronomy community to minimize brightness effects — the same community concern that has followed SpaceX's Starlink from the beginning.
The competitive filing landscape matters. SpaceX has an application for a constellation of up to one million orbital data center satellites. Starcloud, a startup, has filed for up to 88,000. All three proposals plan sun-synchronous dusk-dawn orbits to maximize solar power and all rely on optical inter-satellite links with broadband constellations. None have disclosed detailed satellite specifications or deployment schedules.
The aggregate number across these filings is large enough to matter for orbital congestion. At altitudes between 500 and 1,800 kilometers — a shell already populated by Starlink, OneWeb, and government payloads — adding tens of thousands of objects in co-planar orbital planes creates conjunction risk and coordination complexity that the filings do not address.
The FCC milestone waiver Blue Origin is seeking is the tell. Milestone rules exist to prevent spectrum warehousing: filing for orbital slots without intending to use them, locking up rights while the technology matures. Blue Origin's position is that non-interference access to Ka-band means it is not warehousing in the traditional sense. That may be technically correct. But the effect is the same — 51,600 claimed slots with no committed timeline, competing with SpaceX's million-satellite claim and Starcloud's 88,000. This is the orbital equivalent of buying up domain names and waiting to see which one becomes valuable.
Whether any of these constellations actually get built is a separate question from whether the filings are serious. Blue Origin's FCC filing is a claim on a future state. The hardware does not exist yet at that scale, New Glenn has not demonstrated the launch cadence that would be required, and the economics of space-based compute relative to terrestrial hyperscale facilities remain unproven. But the filing is real, and the spectrum claim is being staked now.
The orbital data center race is not really a technology race. It is a filing race. Blue Origin just entered.

