Cryptocurrency mining is a competition to produce the next block in a blockchain. Miners assemble candidate blocks from pending transactions and hash them repeatedly, changing one field each time, until someone produces a hash below a threshold the network sets. The winner broadcasts the block, everyone else verifies it in milliseconds, and the race restarts.
The work has no value in itself. Its purpose is to make block production costly, so that rewriting history means redoing that cost faster than the rest of the network is producing it. The reward for winning is what pays for the electricity, and the difficulty of winning is what secures the ledger.
What a miner actually does
A Bitcoin miner builds an 80-byte block header containing the version, the previous block's hash, a merkle root summarizing the block's transactions, a timestamp, the current target in compressed form, and a 32-bit nonce. The machine hashes that header with SHA-256 twice and compares the result against the target.
Almost every attempt fails. The nonce field only allows about 4.3 billion values, which a modern machine exhausts in a fraction of a second, so miners also vary an extra field in the coinbase transaction to generate a fresh search space and repeat. There is no shortcut, no way to work out which input will succeed, and no partial credit.
Why the difficulty moves
Bitcoin targets one block roughly every ten minutes. Because total hashrate changes constantly as machines are switched on and off, the network recalculates the target every 2,016 blocks, about two weeks, based on how long the previous 2,016 actually took. Faster than expected, and difficulty rises; slower, and it falls. A single adjustment cannot change difficulty by more than a factor of four in either direction.
This mechanism has an important economic consequence: mining is a zero-sum race for a fixed reward. Adding hashrate does not produce more coins, it divides the same issuance among more machines. Bitcoin's difficulty stood at roughly 127.45 trillion on 8 September 2026, with network hashrate around 930 exahashes per second.
Where the reward comes from
A winning miner collects two things: the block subsidy, which is newly created coin, and the transaction fees paid by the transactions included in the block. The subsidy halves every 210,000 blocks. It has been 3.125 BTC since 20 April 2024, and the next halving is expected at block 1,050,000 in 2028.
Fees are meant to replace the subsidy eventually, and currently do not come close. Fees accounted for roughly 0.43 percent of miner revenue on the day observed in early September 2026, according to reporting on network conditions at the time. The long-run question of what pays for security once the subsidy becomes negligible remains open.
| Period | Block subsidy | Approximate start |
|---|---|---|
| First era | 50 BTC | January 2009 |
| Second era | 25 BTC | November 2012 |
| Third era | 12.5 BTC | July 2016 |
| Fourth era | 6.25 BTC | May 2020 |
| Fifth era | 3.125 BTC | April 2024 |
| Sixth era | 1.5625 BTC | Expected 2028 |
The hardware
Bitcoin mining runs on application-specific integrated circuits, chips that do nothing but compute SHA-256. General-purpose processors and graphics cards were competitive for Bitcoin only in its earliest years; the efficiency gap is now several orders of magnitude and cannot be closed.
The number that matters is efficiency, expressed in joules per terahash. A machine at 16 J/TH produces the same hashrate as one at 32 J/TH for half the electricity, which in a business whose main cost is power decides whether it operates at all. Leading fleets reported efficiency in the mid-teens during 2026; older machines at 30 J/TH and above run only where electricity is very cheap or during price spikes.
Other coins use different algorithms, some of them designed to resist specialized hardware so that graphics cards stay competitive. Those designs have generally slowed the transition to ASICs rather than prevented it.
The economics in three variables
Mining profitability reduces to efficiency, electricity price and hashprice. Hashprice is the revenue the network currently pays per unit of hashrate per day, which combines the coin price, the difficulty and the fee level into a single figure. It was reported at $39.63 per petahash per day in early September 2026, up from $32.42 thirty days earlier.
Given those three numbers, the useful outputs are operating profit per day and break-even electricity price, the tariff at which a machine stops covering its power cost. Comparing break-even against the tariff actually paid answers the only question that matters before buying hardware.
What happens to a block once it is found
The winning miner broadcasts the block to its peers, who check it before passing it on. Verification covers the proof of work itself, the validity of every transaction inside, the block size, and whether the coinbase transaction claims the correct subsidy. A block claiming more than the schedule allows is rejected by every honest node regardless of how much work went into it.
Occasionally two miners find a block at nearly the same time and the network briefly splits, with different nodes seeing different tips. The split resolves at the next block, and the miner whose block was abandoned earns nothing for it. These orphaned or stale blocks are a small but real cost of mining, and they are one reason miners care about how quickly their pool propagates work.
Solo mining and pools
At roughly 930 exahashes per second of network hashrate, a single machine producing 200 terahashes per second represents about one part in 4.6 million of the network. Mining alone, it would expect to find a block roughly once every 88 years, with enormous variance around that average.
Mining pools exist to convert that lottery into an income. Thousands of miners contribute hashrate to a shared effort, and the pool distributes rewards in proportion to the work each contributed. Nearly all mining today happens through pools, which is covered in the pools guide.
