A company backed by nearly half a billion dollars in funding wants to mine Bitcoin in orbit — and critics say it might be the most expensive Bitcoin ever produced. Starcloud, a startup building data centers in space, has drawn skepticism after announcing plans to start mining BTC from orbit, according to a report from Protos published Monday, September 7.
By Michael Nguyen | September 8, 2026
The Hook: Mining Goes Orbital
Starcloud describes itself as a data-centers-in-space company. It has raised close to 500 million USD and carries a valuation above 2 billion USD, per Protos. The company recently said it will begin mining Bitcoin in space — an announcement the outlet called out as doubling down on an already controversial business model.
For regular investors, the question is simple: does putting miners in orbit make any economic sense? Bitcoin mining is, at its core, an energy business. Miners survive by finding the cheapest electricity on Earth — which is why mining farms migrate to hydro dams, flare-gas fields, and wind-rich plains. Space flips that logic on its head.
Why Space Mining Breaks the Economics
Protos lays out the cost stack that Earth-bound miners never have to think about:
- Launch costs — every kilogram of hardware has to be rocketed to orbit, roughly 500 kilometers up.
- Heat is the enemy — in a vacuum, getting rid of waste heat requires large radiators, and heat management is one of the hardest parts of spacecraft design.
- No energy choice — a ground miner can switch from coal to hydro to flared gas. An orbital data center gets whatever sunlight its panels catch and nothing else.
- Maintenance in orbit — even Earth data centers need constant replacement parts. Fixing a server 500 kilometers overhead is a different problem entirely.
- Radiation — chips in orbit take far more radiation damage than chips in a cooled building on the ground.
Protos concludes that Starcloud is pursuing “the most expensive BTC in the history of cryptocurrency.”
The Core Conflict: Experts Say the Whole Model Is Debunked
The criticism goes beyond the mining plan. According to Protos, the Institute of Electrical and Electronics Engineers — IEEE, one of the world’s largest technical professional organizations — has published analysis disputing the feasibility of orbital data centers at scale, and the engineering YouTube channel Real Engineering has done the same.
The numbers cited by Protos are stark. Reaching Starcloud’s stated goal of gigawatts of processing power in space would, per the IEEE critique, take tens of thousands of rocket launches and a combined solar array and radiator cluster roughly four kilometers long and nearly one kilometer wide. For comparison, SpaceX — the most launch-active company on the planet — currently conducts just over one hundred launches per year. The International Space Station, which took 13 years to assemble, spans only about 100 meters by 100 meters.
What Starcloud Has Actually Done So Far
Starcloud is not purely a paper concept. The company has already launched a satellite carrying an Nvidia H100 GPU — one of the most in-demand AI chips on the market. But per Protos, the chip cannot run at full power because the satellite overheats. That detail matters: heat dissipation is the exact problem a mining rig would face, except miners run far hotter and are far less forgiving than a single GPU.
CEO Philip Johnston has promoted the concept in interviews, including a recent appearance with Y Combinator founders and partners. Protos notes the interviewers pressed him with few difficult questions while appearing ready to back the company further.
Market Implications: Why Bitcoin Miners Are Looking Up — Literally
Starcloud’s pivot did not appear in a vacuum. Across the mining industry, companies are chasing the AI compute boom, repurposing sites and power contracts that once hosted Bitcoin rigs. In that environment, attention-grabbing announcements help raise capital — Starcloud’s near half-billion-dollar raise shows investors are willing to fund big swings on compute real estate, even when that real estate is in orbit.
For Bitcoin itself, the story is a reminder of what actually secures the network: earthbound miners competing ruthlessly over electricity prices. Hashrate that costs a fortune to launch and cool does not strengthen Bitcoin’s security budget so much as it burns venture capital. If space mining ever became economical, it would add hashrate — but nothing in the current cost structure suggests that day is near.
There is also a credibility angle for retail holders. Mining stocks have increasingly traded on AI narratives rather than Bitcoin production. When a company valued above 2 billion USD announces orbital mining, investors should ask which part of the story is driving the valuation — the compute, or the headline.
The Verdict
Bitcoin mining in space is currently a marketing-friendly science project, not a business plan. The physics problems — heat, radiation, launches, maintenance — are documented by credible engineering institutions, and Starcloud’s own satellite cannot run its single GPU at full power. Even Protos, which broke down the math, frames the timeline as more than a decade and far more than the half billion raised so far.
Watch what Starcloud does next, not what it says. Another satellite launch with better thermal management would be real progress. Another funding round built on the same pitch would tell you everything else you need to know. Meanwhile, Bitcoin trades near 79,100 USD, ETH around 2,491 USD, and SOL near 104 USD — prices that, for now, are still set by miners and traders firmly on the ground. (Price snapshot: CoinGecko, end of UTC day September 7.)
The cryptocurrency market remains highly volatile. This article is for informational purposes only and does not constitute financial advice.
Disclaimer: This article is for informational purposes only and does not constitute financial advice.
launch costs alone kill the math. even with 500m raised you pay thousands per kg to orbit, that is the most expensive joule in bitcoin mining history and it is not close
^ this. and thats before thermal rejection in vacuum, radiators are heavy af and nobody in the press release mentions them
thousands per kg is actually the optimistic case now, falcon 9 reflight pricing got tighter this year. solar up there beats earth on capacity factor but not by the 10x this needs
most expensive joule in history and people will still buy it because the story sells. remember when solar miners in texas were gonna be free money too
A 2 billion valuation for data centers that do not exist yet, and now orbital mining on top of it. Protos was right to call this one out.
to be fair starcloud said begin mining, not that it would be profitable. still smells like a narrative to justify the raise tho
At a 2 billion valuation every coin they mine up there has to beat terrestrial cost per coin, and the Protos numbers suggest they miss by an order of magnitude. Feels like a raise narrative first, mining second.
the Protos numbers are brutal. even at zero launch cost the orbital payback stretches past the usable life of the hardware, its a raise narrative
order of magnitude feels right. even if starship drops launch costs 10x, the hardware refresh cycle up there is years not months. terrestrial farms just swap ASICs every season
one H100 that overheats in orbit and the next pitch is mining rigs, which run way hotter lol. bold strategy
^ exactly. earth miners chase 2 cent hydro and flare gas, these guys get whatever sunlight the panels catch and nothing else. cant even switch energy sources up there
four kilometers of radiator assembled by robots that dont exist yet, on a launch cadence that doesnt exist yet. love the ambition, hate the physics invoice
Also every repair is a mission with its own launch window. One dead ASIC string up there and your payback calendar is toast.
Exactly. A fuse blows on earth, you swap it the same afternoon. In orbit a dead component means waiting months for the next launch window and hoping the math still works after.
And that assumes the component even fails gracefully. One micrometeoroid hits a coolant line and the whole array is scrap. Terrestrial mining is boring for a reason
A radiator cluster four kilometers long. The ISS took 13 years to assemble and spans 100 meters. The engineering math just is not there.
4km of radiators also means station keeping across that whole span. one thruster misbehaves and your array tumbles into itself. hard pass on the physics
the ISS comparison undersells it honestly, that station was never built to shed megawatts. still agree the radiator mass budget is where this whole plan dies quietly
the radiator mass math is brutal, every megawatt of waste heat up there needs tons of radiator surface. somehow that slide never makes the investor deck
the radiator slide does show up actually, they claim 4kg per kW with deployable panels. still skeptical but at least read the technical paper before dunking
4kg per kW with deployables is their number, independent mass budgets usually land 2-3x that once you add structure and margins. still the only defensible slide they have
right, and at 2k/kg to orbit even a perfect radiator pays for a lot of ASICs down here. the 500m raise buys maybe one demo module and a lot of press releases
capacity factor is the honest argument and nobody makes it. uninterrupted solar beats any terrestrial farm until you price the radiator mass per megawatt
counterpoint: nobody funds a 500m raise without a payload customer lined up. starcloud probably sells compute first and mines btc with the leftover power as a demo. the mining part is marketing
compute first is the only version that pencils. btc mining on leftover orbital power is a press release, the H100 contracts are the business
compute first makes way more sense than mining. selling orbital H100 hours to labs with too much grant money, btc is just the idle load PR story