Orbital has taken its first regulatory step toward an ambitious space computing network. The five-month-old startup has asked the Federal Communications Commission (FCC) for permission to deploy up to 100,000 data center satellites, with the goal of delivering 10 gigawatts of computing power to meet rising demand for artificial intelligence.
The application, filed on June 24, reveals new technical details about the proposed constellation, including the size of the satellites, their planned orbit and how they would exchange data. It comes just weeks after the Los Angeles-based company emerged from stealth with $5 million in pre-seed funding.
Orbital is entering a niche that is beginning to attract more attention across the space industry. As SpaceNews reports, several companies are now exploring orbital data centers as growing demand for AI pushes conventional facilities to their limits in terms of power, cooling and available land.
A Constellation Built To Handle Computing Workloads
Orbital‘s proposal centers on 100-kilowatt-class satellites operating in low Earth orbit between 500 and 850 kilometers above Earth. Each spacecraft would have solar arrays and radiators stretching roughly 100 meters, while its dry mass would range from 1.5 to 2.5 metric tons.
The satellites are designed to process data rather than provide communications services. Instead of relying on their own communications network, they would use optical intersatellite links connected to third-party constellations. The FCC filing, cited by SpaceNews, identifies SpaceX’s Starlink as one example of the systems that could support those data connections.
The company is preparing a demonstration mission for next year, although the first payload will be much smaller than the hardware planned for future operational satellites.
“The demonstrating payload is going to be a very, very scaled down version of what we’re looking to do with a single GPU,” founder and CEO Euwyn Poon said. “Maybe one one hundredth the size.”
First Operational Satellite Targeted For 2028
Orbital plans to follow the demonstration mission with Orbital-1, its first dedicated compute satellite, currently slated for 2028. Poon said the spacecraft is being designed to resemble the future production version as closely as possible, even if the full constellation is not expected to be deployed until well into the next decade.
Some specifications could still change before the design is finalized. Poon noted that Starcloud is targeting 200-kilowatt satellites for its proposed 88,000-spacecraft constellation. SpaceX has also outlined 150-kilowatt-class orbital data centers after filing plans for up to one million spacecraft.
Orbital’s team currently consists of six people with experience at SpaceX, Amazon and Northrop Grumman, combining expertise in spacecraft engineering, large-scale manufacturing and technology development.
Manufacturing And Launch Are Key To The Project
Poon said building satellites at scale will be one of the company’s biggest challenges. In his view, orbital data centers are relatively straightforward systems built around solar panels, radiators and electronics, although they must also operate in the vacuum of space and withstand radiation.
“The complexity is all launch,” Poon said, pointing to the importance of heavy-lift launch vehicles for deploying constellations of this size.
SpaceNews also notes that Orbital is among several companies waiting for SpaceX’s Starship to become available before large orbital computing networks can realistically be deployed.
While the core satellite platform is being designed in Los Angeles, Orbital is also looking for manufacturing partners and broader collaboration opportunities. Poon compared the effort to his experience leading Spin, the electric scooter company he founded before selling it to Ford. He said relatively small design changes, such as switching to swappable batteries, significantly improved operations at Spin, and believes similar refinements could have an equally large impact when applied across a fleet of 100,000 satellites.
