The headline hit my feed at 3:14 a.m. Kuala Lumpur time. A company called Starcloud wants to mine Bitcoin from orbit. The copy said the plan "could revolutionize energy efficiency and reshape the economics of the industry." No launch date. No payload mass. No wattage. No cost per coin. No named launch provider. Just a company, a concept, and a countdown clock that does not exist.
I have chased this exact ghost before. Chasing the green candle through the fog of 2017 taught me one rule that has never failed me: when a narrative arrives before a spreadsheet, the spreadsheet almost never comes. The story is the product. The physics is an afterthought.
So let me do what the original item refused to do. Let me open the machine — sunlight, heat, radiation, and the rocket bill — and show you why "orbit" is not a technical breakthrough. It is a location. And locations are priced, not praised.
Why this story exists now
Space and crypto have been dating on and off for a decade, and the relationship always follows the same beat. Blockstream put Bitcoin over satellite years ago — but that was a broadcast problem, not a mining problem, and it worked precisely because radio is cheap and the network is small. SpaceX has flown crypto payloads more than once. The orbital data center as a concept got a second life in 2024 and 2025, mostly because AI's hunger for compute collided with Earth's power grid, its land-use fights, and its grid-connection queues that now stretch for years in some markets. Starcloud is riding that collision. It is not creating it.
Here is the part most coverage skips, and it matters more than any engineering detail. Starcloud is not a protocol. It is not a DAO. There is no token, no governance vote, no liquidity pool, no airdrop, no vesting cliff. It is a private company that mines BTC and books the revenue, with shareholders capturing value through a traditional cap table. That single fact rewrites how you should read the news. There is no unlock schedule to model. There is no total value locked to track. There is no emissions curve to front-run. There is a profit-and-loss statement, and right now that statement is a blank page with a nice logo and a launch window it will not name.
This is why the item reads as narrative infrastructure rather than a data-driven report. Two data points were actually available: a company plans to launch a satellite to mine Bitcoin, and this might reshape the industry's energy economics. Everything else is inference. I will flag my inferences as inferences. The physics, though, I will state as physics, because physics does not care about your funding round.
The sunlight argument is half true
Lead with the good news, because there is some. At Earth's orbital distance, the solar constant is roughly 1,361 watts per square meter. No clouds. No weather. No atmosphere stealing photons before they reach the panel. A solar array in the right orbit harvests more usable energy per square meter than almost anything you can build on the ground, and it never pays an electricity tariff. That is the entire bull case, and it holds right up until you ask where the satellite actually is.
Low Earth Orbit is not a sunny beach. A satellite there loops the planet in about ninety minutes, and it spends a large slice of that — call it thirty to forty-five minutes — in Earth's shadow. The real sunlight duty cycle in LEO lands somewhere around 60 to 70 percent. You only approach continuous illumination in a very specific geometry: a sun-synchronous orbit near the terminator line, where the satellite rides the day-night boundary and effectively never sets. So "24-hour uninterruptible solar power" is true only inside a boutique orbital slot. Everywhere else you are running a battery, and a battery is dead weight you paid a rocket to lift and will pay again to keep alive. The claim is not false. It is narrow, and narrow claims are the ones you can audit.
The heat problem nobody mentions
Here is where the dream usually dies, and the original item did not mention it once.
Space is not cold the way people imagine. Space is empty, and empty is a terrible conductor. There is no air to carry heat away by convection, no atmosphere to dump into. Your only escape route is thermal radiation, and radiation follows the Stefan-Boltzmann relationship — radiated power scales with the fourth power of temperature. Run the numbers at a sensible operating temperature of around 300 Kelvin and you are radiating roughly 400 to 800 watts per square meter of surface. That is the ceiling, and it is brutally low.
Now park an ASIC next to it. A modern machine like an S21 pulls down around 3.5 kilowatts. Divide one by the other and you need several square meters of radiator just to keep one miner from cooking itself — and that radiator needs a matching array of solar panels to feed it. The satellite platform carrying a single mining rig becomes a structure the size of a small room, most of which is plumbing for heat, not hashing power. The miner is a passenger. The radiator is the vehicle.
This is the inversion the marketing hides. On the ground, cooling is an operating cost — you build a warehouse, you move air or fluid, you pay for it slowly over years. In orbit, cooling is a launch cost. You pay for every square meter of radiator at liftoff, in kilograms, in the most expensive shipping lane humans have ever built. The heat did not disappear. It got a boarding pass.
Radiation is a silent tax on your hashrate
Third constraint, and the quietest. Space is full of energetic particles. When a cosmic ray punches through a silicon die, it can flip a bit — a single-event upset. Your ASIC is not rad-hardened. It was designed for a climate-controlled room in Texas or Kazakhstan, not for a particle bath above the atmosphere.
You get two options, and both cost you. Wrap the chip in shielding and error correction, which adds mass and complexity and cost. Or accept a higher silent failure rate, which means some fraction of your nameplate hashrate is fiction. There is no third door. The effective hashrate of an orbital miner is always smaller than its advertised number, and the gap is invisible on a marketing slide. On the ground, a bad board throws an error and a technician swaps it before lunch. In orbit, the board is gone. You cannot send a truck. You cannot send a hand. Write this down: in orbit, every hardware fault is a total loss, because there is no maintenance — only deorbit. A ground farm upgrades. An orbital farm slowly becomes debris with a depreciation schedule.
Then add the boring constraints. Orbital debris tracking and collision-avoidance maneuvers cost fuel and attention. Communications bandwidth to a ground station limits how fast you can even report your own hashrate, which matters if you are selling a live-performance story. Insurance for a mining payload is either unavailable or priced like a lottery ticket, because underwriters have no actuarial history for "Bitcoin ASIC in vacuum." None of this appears in the pitch. All of it appears on the invoice.
The only number that matters
Now the economics, where the whole thing lives or dies.
The energy is nearly free at the margin. Granted. But orbital mining does not compete on energy cost. It competes on total cost of ownership, and TCO is dominated by one variable the company does not control: the price per kilogram to orbit. Current commercial launch sits somewhere between 1,500 and 6,000 dollars per kilogram depending on vehicle and rideshare terms. The dream case is a fully reusable heavy-lift stack pushing that toward 100 dollars per kilogram.
Do the arithmetic. One 3.5-kilowatt miner, plus radiator, plus solar array, plus structure, plus comms, plus platform overhead — call it a few hundred kilograms of dead metal just to support a single rig. At 1,500 dollars per kilogram you have spent hundreds of thousands of dollars to deploy a machine that on Earth costs a few thousand and earns perhaps ten to fifteen dollars a day at current difficulty. That payback horizon does not belong in an investor deck. It belongs in a horror novel. At 100 dollars per kilogram the equation flips — suddenly the radiator is cheap, redundancy is affordable, and orbital sunlight becomes a genuine arbitrage against grid power. So the real bet is not on Starcloud. The real bet is on the launch cost curve. Everything else is downstream of that one line.
And note what this does to the competition. Ground miners like Marathon, Riot, and CleanSpark have cheap power contracts, mature supply chains, and — critically — the ability to repair, upgrade, and resell. They are not disrupted by a single experimental satellite. They are, at most, briefly distracted by it while investors chase the shiny object.
My own field notes
I do not write this from a textbook. I write it from having watched a version of this movie before.
In 2020, during DeFi Summer, I flagged a yield-farming flaw without reading a line of Solidity. I sat in Discord channels and watched how users talked, and the "yield bleed" was obvious in the nervous rhythm of the conversation. The code was confident; the crowd was not. That gap is the trade. This Starcloud story has the same shape — a confident narrative sitting on top of an absent dataset. The temperature of the conversation is hot. The information is cold.
In 2025, testing an AI trading agent through a live volatility spike, I watched the bot panic on social-media noise. It overreacted to sentiment and underweighted structure. The developers missed it because they only read the logs. I caught it because I watched the tape while the market breathed. Same lesson here, different arena. Do not read the press release. Watch the engineering. The press release is the noise; the physics is the signal. Speed is the only asset that never depreciates — and speed without verification is just noise with a deadline.
The contrarian cut: the winners are not the miners
Everyone is watching Starcloud and asking whether it can mine Bitcoin. Wrong question.
The clean winners of any verified "orbital compute" model are upstream: launch providers, satellite bus manufacturers, solar array and radiator suppliers, rad-hard electronics vendors. If a mining satellite is ever proven, it does not threaten Marathon or Riot or CleanSpark — those operations have cheap power contracts and, crucially, they can be repaired. It does one thing: it creates a brand-new category of payload, the compute satellite, and hands a recurring order book to the companies that build and launch them. That is the asymmetry. The mining company carries all the risk and captures a thin spread. The launch company carries none of the mining risk and books the fare either way.
There is a second blind spot, and it is regulatory. If a satellite mines Bitcoin above the ocean, in international airspace, or parked over a specific nation, which jurisdiction taxes it? Which regulator supervises it? The Outer Space Treaty of 1967 places continuous authorization and supervision of a non-governmental entity's activities on the launching state — so a US launch means FAA and FCC licensing, ITU coordination for spectrum and orbital slot, and ITAR or EAR controls if the satellite tech crosses borders. But the coin itself sits in a legal vacuum. No precedent exists for the tax residency of a Bitcoin minted in orbit. That is not a footnote. That is a future test case, and it is entirely unpriced.
Where I land, and what I am watching next
This is a narrative-first item, and in a bear market narratives are how capital distracts itself from the bleeding. Liquidity vanishes faster than a dream in DeFi, and narrative liquidity behaves the same way — it pools fast and drains faster. "Space plus crypto" is a cyclical guest, not a new permanent resident. Historically it has a short half-life, usually three to six months, unless a real milestone arrives to reset the clock.
So here is what I will actually watch, in order. First, did the satellite reach orbit and power up — a verifiable yes or no. Second, is the sustained hashrate reported in terahashes, or in adjectives. Third, what is the disclosed cost per Bitcoin, benchmarked against the ground-miner average, because that single figure decides whether this is a business or a billboard. Fourth, where does the launch price curve go, because that is the true underlying asset and the one thing that can reprice the whole category overnight. And fifth, does any tax authority or securities regulator plant a flag on the jurisdiction question, because the first ruling will define the rules for everyone who follows.
Fifty percent down, one hundred percent ready. I am not short the idea. I am short the story that skips the numbers. Show me the radiator, show me the effective hashrate after radiation, show me the cost per kilogram, and I will reprice everything I just wrote. Until then, this is a lovely orbit and an empty spreadsheet.
The candle is green. The fog is thick. Watch the tape — and this time, watch the launch pad too.