Starcloud Space Bitcoin Mining: Why an Nvidia-Backed Startup Is Sending ASICs to Orbit
A 60-kilogram satellite carrying an Nvidia H100 GPU is already running large language models 325 kilometers above Earth. Now the company behind it wants to mine Bitcoin in space — and it just filed paperwork for 88,000 more satellites to follow.
From Science Fiction to Starcloud-1
Starcloud, formerly known as Lumen Orbit and backed by Nvidia, launched its first satellite on a SpaceX rocket in November 2025. The Starcloud-1 spacecraft carried the most powerful GPU ever sent to orbit — an Nvidia H100 chip roughly 100 times more capable than any prior space-based processor. Within weeks, the satellite was running Google's Gemma large language model and, by December, became the first spacecraft to train an LLM in orbit using Andrej Karpathy's nanoGPT.
That demonstration proved the fundamental thesis: serious computation can happen outside Earth's atmosphere. Now CEO Philip Johnston is pivoting from proof-of-concept AI workloads to something far more provocative — Bitcoin mining from low Earth orbit.
The Starcloud-2 Mission: ASICs Among the Stars
Starcloud-2, scheduled for launch in late 2026, will carry a cluster of application-specific integrated circuit (ASIC) miners alongside GPUs. If the mission succeeds, it will produce the first Bitcoin ever mined off-planet.
The economic logic is straightforward: ASICs are purpose-built for SHA-256 hashing and dramatically cheaper per kilowatt than GPUs. Johnston estimates that a 1-kilowatt ASIC costs roughly $1,000, compared to about $30,000 for a 1-kilowatt Nvidia B200. That 30-to-1 cost advantage per unit of computing power makes ASICs the natural candidate for orbital deployment, where every kilogram of payload matters.
But hardware economics alone don't explain why mining in space might work. The real advantage lies in two resources that are virtually free in orbit: energy and cooling.
The Physics of Orbital Mining
Unlimited Solar, Zero Cooling Bills
In low Earth orbit, satellites receive near-continuous sunlight. There is no atmosphere to scatter photons, no clouds to block panels, and no nighttime — at least not in the prolonged sense experienced on the ground. Solar panels in orbit can generate power at efficiencies that Earth-based installations cannot match.
Cooling is the other half of the equation. Bitcoin mining generates enormous heat: every watt of electricity consumed by an ASIC becomes a watt of thermal energy that must be dissipated. On Earth, miners spend billions on liquid cooling systems, industrial fans, and climate-controlled facilities. In orbit, the vacuum of space provides passive thermal radiation at temperatures approaching minus 270 degrees Celsius. No pumps, no coolant, no air conditioning — just radiative heat transfer into the cosmic background.
Johnston claims these twin advantages could reduce energy costs by a factor of ten compared to terrestrial operations.
The Numbers on Earth
To appreciate what "10x cheaper" means, consider the current baseline. In 2026, mining a single Bitcoin requires approximately 854,400 kilowatt-hours of electricity — enough to power an average U.S. household for over 81 years. At the industry-standard rate of $0.06 per kWh, that translates to roughly $51,000 in electricity per coin. Miners operating at higher rates, or with less efficient hardware, face even steeper costs.
Meanwhile, Bitcoin's network hashrate has surged past 894 EH/s, briefly touching 1 ZH/s (1,000 EH/s) in January 2026 before pulling back. The April 2024 halving cut block rewards to 3.125 BTC, squeezing margins to the point where hash price has dropped below the $35-per-petahash-per-second-per-day break-even threshold. Only miners with electricity below $0.06/kWh and hardware efficiency under 20 joules per terahash are surviving.
If space-based solar genuinely delivers energy at one-tenth the terrestrial cost, orbital mining could undercut even the cheapest hydro-powered facilities in Paraguay or Iceland.
The Launch Cost Problem
The promise of free energy must be weighed against what remains the most formidable barrier: getting hardware to orbit.
SpaceX's Falcon 9 rideshare program currently charges $350,000 for up to 50 kilograms to sun-synchronous orbit, with each additional kilogram priced at $7,000. A meaningful mining operation — say, 500 kilograms of ASICs plus power systems, thermal radiators, and communication hardware — could easily cost several million dollars in launch fees alone.
Those numbers change dramatically if SpaceX's Starship reaches full operational capability. With projected costs of $13 to $32 per kilogram under high-reusability scenarios, Starship would reduce the launch bill for that same 500-kilogram payload from millions of dollars to under $16,000. Whether those projections materialize on the timeline Starcloud needs is an open question, but the trajectory of launch costs is unmistakably downward.
This is precisely why Starcloud filed an application with the U.S. Federal Communications Commission in February 2026 for a constellation of up to 88,000 satellites. The vision is not a single mining rig in orbit but a distributed orbital data center network capable of running AI workloads and blockchain computation at planetary scale.
Mining Hardware's Shelf Life Problem
Even if launch costs plummet, orbital mining faces a challenge unique to the Bitcoin industry: hardware obsolescence.
ASIC miners follow a relentless efficiency curve. Each new generation delivers more terahashes per watt, rendering older models unprofitable as network difficulty adjusts upward. On Earth, a miner can swap out machines every 12 to 18 months. In orbit, there is no maintenance crew, no hardware refresh, and no returns. Once deployed, a satellite's ASICs must compete against whatever the industry produces on the ground over the satellite's operational lifetime.
Starcloud-1 has an expected mission lifetime of 11 months at its 325-kilometer altitude — roughly one ASIC generation cycle. If orbital mining is to be more than a novelty, future satellites will need to operate at higher altitudes with longer lifespans, which introduces additional radiation exposure and communication latency challenges.
Johnston acknowledges this tension directly, noting that profitability can erode rapidly as new hardware reaches the market. The counter-argument is that if energy costs are sufficiently low, even slightly outdated ASICs remain profitable longer than they would on Earth.
The Orbital Data Center Race
Starcloud is not operating in a vacuum — metaphorically speaking. The orbital data center market, valued at roughly $1.77 billion by 2029, is projected to reach $39 billion by 2035 at a 67% compound annual growth rate. Eight organizations filed plans, launched hardware, or committed funding for orbital compute within the last 90 days alone.
Key competitors include:
- OrbitsEdge, partnering with Hewlett Packard Enterprise for AI-enabled space experiments, with a first orbital demonstration slated for 2026.
- Lonestar Data Holdings, pursuing a dual LEO-and-lunar strategy with first commercial LEO service targeted for Q4 2026.
- Aetherflux, developing orbital compute nodes with a first operational target of Q1 2027.
- Google and major cloud providers, each advancing along distinct technical paths toward space-based computing.
None of these competitors have publicly announced Bitcoin mining ambitions. Starcloud's willingness to combine AI workloads with blockchain computation may prove to be either a visionary multi-revenue strategy or a distraction from the core orbital computing business.
What Success Would Mean for Bitcoin
If Starcloud-2 successfully mines even a fraction of a Bitcoin in orbit, the implications extend beyond a single company's balance sheet.
Decentralization gains a new dimension. Today, Bitcoin mining concentrates wherever electricity is cheapest — hydroelectric dams in Sichuan, geothermal plants in Iceland, natural gas flares in Texas. Orbital mining would add a location that no government controls and no grid operator can shut down. For a network built on censorship resistance, space-based hashrate represents the ultimate jurisdictional diversification.
The energy narrative shifts. Bitcoin's environmental critics focus on fossil fuel consumption. Mining powered entirely by solar panels in orbit eliminates carbon emissions from the equation — though critics would rightly point to the carbon footprint of rocket launches themselves.
A new mining economics tier emerges. If orbital energy costs prove as low as claimed, space-based miners could operate profitably even during severe market downturns, creating a hashrate floor that makes the network more resilient to price-driven miner capitulation.
The Sober Reality
For all its appeal, orbital Bitcoin mining remains deeply experimental. The Starcloud-2 mission has not launched. No Bitcoin has been mined in space. The economics depend on assumptions about solar panel efficiency, thermal management, ASIC longevity in radiation environments, and communication bandwidth for submitting blocks to the network — none of which have been validated at mining scale.
The 88,000-satellite constellation exists only as an FCC filing. Even Starlink, with the full resources of SpaceX behind it, took years to deploy thousands of satellites. Building an orbital mining network of that scale would require capital, manufacturing capacity, and regulatory approvals that dwarf what any crypto-native company has achieved.
And there is a deeper question: by the time orbital mining infrastructure reaches economic viability, will Bitcoin's block rewards — halving again around 2028 — still justify the investment? Or will the transaction fee market need to mature dramatically for space-based mining to pencil out?
Looking Up
Starcloud's bet is ultimately not about Bitcoin mining alone. It is about proving that computation in orbit is economically viable for any workload — AI training, inference, data processing, and yes, cryptocurrency mining. Bitcoin happens to be a convenient first use case because ASICs are cheap, energy-hungry, and produce a directly monetizable output that can be transmitted back to Earth as a signed transaction.
The next 12 months will reveal whether orbital mining moves from headline to hashrate. If Starcloud-2 launches on schedule and successfully mines Bitcoin from space, it will mark a inflection point not just for crypto but for the entire economics of off-planet computing.
The race to mine Bitcoin has always followed the cheapest energy on Earth. In 2026, that race might finally leave the planet.
Building decentralized infrastructure demands reliable foundations — whether on Earth or in orbit. BlockEden.xyz provides enterprise-grade RPC endpoints and API services for leading blockchains including Ethereum, Sui, and Aptos. Explore our API marketplace to power your next project.