Blue Origin's Project Sunrise: The Race to Build Data Centres in Orbit
Data Infrastructure / Infrastructure Management | 6 min read
Blue Origin, Jeff Bezos' space company, has filed an official application with the US Federal Communications Commission (FCC) to launch and operate a constellation of up to 51,600 satellites under a programme called Project Sunrise. The filing, submitted on 19 March 2026, positions the orbital network as a new tier of AI computing infrastructure — designed to operate independently of Earth's land, power, and cooling constraints. The proposal marks Blue Origin's most ambitious move to date, entering a rapidly forming race for space-based compute alongside SpaceX, Starcloud, and Google's Project Suncatcher.
What Project Sunrise Actually Proposes
Project Sunrise would consist of up to 51,600 satellites operating in circular, sun-synchronous orbits at altitudes between 500 and 1,800 kilometres, with each orbital plane containing approximately 300 to 1,000 satellites. The constellation satellites would feature at least three antenna variants to address a range of coverage requirements, and would transmit data primarily through optical inter-satellite laser links. Traffic would be routed through Blue Origin's separate TeraWave constellation — a planned 5,408-satellite ultra-high-speed network that would serve as the communications backbone — and through other mesh networks to connect with ground stations.
"The built-in efficiencies of solar-powered satellites, always-on solar energy, lack of land or displacement costs, and nonexistent grid infrastructure disparities fundamentally lower the marginal cost of compute capacity compared to terrestrial alternatives."
— Blue Origin FCC Filing, Project Sunrise
Why the Proposal Makes Sense Now
The logic behind orbital data centres is grounded in a genuine terrestrial infrastructure crisis. The International Energy Agency reported that data centre and AI-related electricity consumption — roughly 460 terawatt-hours in 2022 — was expected to more than double to over 1,000 terawatt-hours by 2026, an amount roughly equivalent to Japan's entire national energy consumption. Land is scarce in key markets, electrical grids are strained, and water supplies used for cooling are under pressure in drought-prone regions. Blue Origin directly addresses this in its FCC filing, stating that Project Sunrise will "ease mounting pressure on US communities and natural resources by shifting energy- and water-intensive compute away from terrestrial data centres, reducing demand on land, water supplies and electrical grids."
Sun-synchronous orbits are central to the energy strategy. Unlike terrestrial data centres that depend on grid connections or on-site generation, Project Sunrise satellites would operate using continuous solar power — avoiding the interruptions caused by night-time and weather that hamper ground-based solar. The orbital environment eliminates land-use costs and removes grid infrastructure constraints entirely, which Blue Origin argues fundamentally lowers the marginal cost of AI compute capacity at scale.
A Crowded Race: SpaceX, Starcloud, and Google
Project Sunrise enters a rapidly accelerating field. In January 2026, SpaceX filed with the FCC for authorisation to deploy up to one million orbital data centre satellites — a proposal Amazon called on the FCC to deny, arguing it would take "centuries" to deploy. Starcloud, a startup focused on orbital data centres, has proposed a constellation of up to 88,000 satellites. Google's Project Suncatcher — a long-term "moonshot" — aims to come online by 2027 via partner Planet Labs. Nvidia announced earlier in March 2026 that it would begin producing space-optimised versions of its most advanced processors, including those used for AI training and inference, with CEO Jensen Huang declaring: "Space computing, the final frontier, has arrived."
Practical Hurdles and Regulatory Strategy
Despite the scale and ambition of the proposal, significant practical challenges remain. Blue Origin has only flown its New Glenn rocket twice, and deploying tens of thousands of satellites would require sustained launch capacity over many years. The TeraWave backbone constellation has yet to launch a single satellite, with the first TeraWave orbiters not expected before end of 2027. In its FCC filing, Blue Origin has asked for a waiver from milestone rules that would typically require half the constellation to be in orbit within six years of approval and the full constellation within nine. The company accepts subordinate spectrum status in exchange for reduced regulatory friction — a positioning move that establishes its place in the queue of similar filings.
Blue Origin commits to following guidelines to limit orbital debris and plans to deorbit satellites within five years of the end of their operational life, while also working with the astronomy community to reduce the constellation's visual impact on the night sky.
Key Takeaways
- • Blue Origin filed with the FCC on 19 March 2026 to deploy up to 51,600 satellites under Project Sunrise — a distributed orbital data centre constellation designed for AI and compute workloads.
- • Satellites would operate in sun-synchronous orbits at 500–1,800km, connected via laser inter-satellite links and routed through the TeraWave broadband backbone — and powered entirely by continuous solar energy.
- • The proposal is driven by a genuine terrestrial infrastructure crisis: global data centre electricity demand is expected to exceed 1,000 TWh by 2026 — roughly equivalent to Japan's entire national energy consumption.
- • Blue Origin joins SpaceX (1M satellites), Starcloud (88,000 satellites), and Google's Project Suncatcher in a fast-forming race for orbital compute — with Nvidia producing space-optimised AI processors to support it.
- • Key hurdles remain: New Glenn has only flown twice, TeraWave has yet to launch, and Blue Origin has requested an FCC waiver from deployment milestones. Project Sunrise is a regulatory positioning move as much as an immediate operational plan.