Bitcoin's Next Difficulty Retarget Pushes Older Miners Toward a 3-Cent Power Bill
A projected difficulty increase cuts the break-even electricity rate for every rig on the network, even with Bitcoin's price and block rewards flat. The math is unforgiving for older machines and anyone paying retail for power. Here's who survives the next two retargets.
The most important number in Bitcoin mining this month isn't the price of Bitcoin. It's a difficulty readout that most miners won't bother checking until it's already cost them money.
Here's the setup. Difficulty adjusts every 2,016 blocks, roughly once every two weeks, and the current projection points to another increase at the next retarget. That means every machine on the network has to work harder for the same block reward. Nothing about the reward changes. Nothing about the price changes. The only variable that moves is how much hash you need to produce to earn 3.125 BTC per block, the reward set after the April 2024 halving.
Here's what matters: the adjustment lowers the electricity rate a less efficient rig can afford. Not in a year. At the next retarget.
What The Retarget Actually Does
Start with hashprice, the revenue a unit of hashrate earns per day. With Bitcoin near $90,000 and network hashrate in the high 700s exahash, that number has been sitting somewhere around $45 per petahash per day, fees included. It moves constantly. Difficulty is the lever that pushes it down mechanically, and it doesn't care what your thesis is.
Now take a mid-tier rig running at 25 joules per terahash. That machine pulls 0.6 kilowatt-hours per terahash per day. At $45 per petahash, each terahash earns about 4.5 cents. Divide one by the other and break-even power lands near 7.5 cents per kWh. Anything above that's margin. Anything below it's a subsidy.
Roll the same math forward with a 5% difficulty increase. Revenue per terahash drops to roughly 4.3 cents. Break-even falls to about 7.1 cents. A 0.4 cent move sounds like noise. On a 30 megawatt facility running 30,000 machines, it's a real line item on the power bill every hour of every day.
Then run the older hardware. A 60 J/TH machine from the last cycle draws 1.44 kWh per terahash per day for the same 4.5 cents of revenue. Break-even power cost lands near 3 cents per kWh. There are only a handful of places on earth where 3 cent power exists at industrial scale, and most of them come with a gas flare, a dam, or a very long transmission line attached.
The numbers tell the story. Efficiency is the whole game, and the network keeps raising the bar on what counts as efficient.
Who Wins, Who Gets Repriced
So who's fine? The operators running 15 to 18 J/TH fleets behind long-term power contracts under 4 cents per kWh. A 5% retarget barely registers for them. What registers is the megawatt price on a ten-year agreement signed before the crowd showed up.
Who gets repriced? Anyone paying retail. Anyone hosted at a facility that passes costs straight through. And anyone still running S19-class hardware bought near 2021 prices, waiting for a cycle to bail them out. Frankly, that last group always gets hurt first, and they're the least likely to have modeled the retarget at all.
The public miners that report all-in power costs in the $0.03 to $0.045 range per kWh aren't sweating this. The ones reporting $0.06 and up are already running the shutdown math on their worst-performing racks, whether they've said so or not.
Here's my first strong take. Bitcoin mining stopped being a Bitcoin business a while ago. It's a power business with a software wrapper. The winners own generation, negotiate interruptible load deals, and treat hashrate as a byproduct of energy arbitrage. The losers are the ones who think they're trading Bitcoin with extra steps. If you can't describe your operation in cents per kilowatt-hour, you don't have a mining thesis. you've exposure.
Second take, and this one matters for price action. Difficulty compression forces treasury selling. When hashprice falls and margins thin, the marginal miner doesn't shut off overnight. It sells the BTC it mined last week to cover payroll this week. That's structural sell pressure that never shows up in ETF flow data and never gets attributed to its real cause. Watch miner outflows to exchanges after each retarget. That's where the tell is.
And the uncomfortable question: if the projected adjustment means a chunk of the fleet can't earn its own electricity, why is total hashrate still climbing? Because someone out there's comfortable at these margins, and they're taking share from the operator paying 9 cents. Difficulty isn't a random walk. It's a competitive purge, and it runs every two weeks.
The Takeaway
Track two numbers instead of one. Hashprice in dollars per petahash per day, and your own break-even power rate. If the second number is higher than the first, the machine is a losing position and shutting it down is the trade.
From a risk perspective, the retarget is the cleanest signal in the asset class. It tells you exactly how much the network is willing to pay for hashrate, and it updates on a schedule nobody can lobby, front-run, or talk their way out of.
The miners who make it through the next two adjustments won't be the ones who called Bitcoin's price right. They'll be the ones who locked in cheap power before anyone else bothered to run the math.
Explore More
Key Terms Explained
Profiting from price differences of the same asset across different markets.
The first cryptocurrency, created in 2009 by the pseudonymous Satoshi Nakamoto.
A bundle of transactions that gets permanently added to the blockchain.
The cryptocurrency given to miners or validators for successfully adding a new block to the blockchain.