Startup Starcloud Plans First Bitcoin Mining Satellite in Low-Earth Orbit

The company first proved its concept in November 2025 when it launched Starcloud-1 aboard a SpaceX Falcon 9 rocket. The small, refrigerator-sized satellite carried five Nvidia H100 processors, marking the first time data-center-grade GPUs operated in orbit.

During that mission, Starcloud successfully trained a small large language model in space and ran inference from a version of Google Gemini. The test demonstrated that powerful computing workloads—traditionally housed in massive terrestrial facilities—can function in the vacuum of space.

Now the company is preparing a larger follow-up satellite. CEO Philip Johnston said Starcloud-2, expected to launch later in 2026, will carry a much larger GPU cluster along with specialized bitcoin-mining hardware known as ASICs (Application-Specific Integrated Circuits).

Starcloud argues these conditions could dramatically lower energy costs. The company estimates orbital data centers could operate with energy costs roughly 10 times lower than traditional facilities while also delivering similar reductions in lifetime carbon emissions.

To support that vision, the startup has filed paperwork with the Federal Communications Commission (FCC) seeking approval for a constellation of as many as 88,000 satellites dedicated to orbital computing infrastructure. Johnston has outlined a long-term goal of constructing a 5-gigawatt orbital data center supported by massive solar arrays stretching several kilometers.

ASIC mining machines often cost hundreds or a few thousand dollars per unit, compared with tens of thousands for advanced AI GPUs. Because mining primarily converts electricity into cryptographic computations, the availability of constant solar energy makes it a natural fit for off-planet power generation.

Falling launch costs are helping revive interest in the concept. Modern rockets are steadily reducing the cost per kilogram of payload delivered to orbit, making it more feasible to deploy computing hardware and large solar arrays above Earth.

At the same time, improvements in mining hardware efficiency have slowed compared with earlier years. That longer replacement cycle means ASICs could remain useful for longer periods, potentially improving the economics of running them in space.

Despite the excitement, the approach faces significant hurdles. Space hardware must survive radiation exposure, temperature swings, and potential collisions with debris. Satellite reliability also remains a concern—one of Starcloud-1’s GPUs reportedly failed before launch.

Regulators and environmental groups are also watching closely. Large constellations raise questions about orbital congestion and the risk of cascading debris events, sometimes referred to as the Kessler syndrome.

Even within the technology sector, some observers remain skeptical about near-term feasibility. Critics argue that maintaining high-performance computing systems in orbit introduces logistical challenges that could outweigh the energy benefits, at least in the early years.

Still, interest in orbital computing is spreading quickly. Companies including SpaceX, Blue Origin, and Axiom Space have all outlined plans involving orbital infrastructure that could support computing workloads.

Market forecasts suggest orbital data-center services could grow rapidly over the next decade, potentially reaching tens of billions of dollars as satellite manufacturing and launch technology mature.