Proof of Work (PoW): How Miners Secure a Blockchain

Proof of Work (PoW): How Miners Secure a Blockchain

Proof of work (PoW) is a consensus mechanism, a rule that lets many computers agree on one shared record without a central party, in which participants show costly computing effort that anyone can check quickly. In Bitcoin, the effort is finding a block’s hash at or below a set target.

Proof of Work Began as a Defense against Junk Mail

The idea behind proof of work was proposed to fight junk mail years before Bitcoin existed. In a paper titled “Pricing via Processing or Combatting Junk Mail,” presented in 1992 and published in 1993, Cynthia Dwork and Moni Naor suggested that a user should have to perform a moderately hard computation to gain access to a shared resource. The cost was meant to discourage frivolous use, and the authors aimed in particular at junk mail.

In general, a proof of work is data that is costly and time-consuming to produce but easy for others to verify. Hashcash, a system created by Adam Back, attached a hash-based proof of this kind to each email to deter spam and denial-of-service abuse. Denial of service means flooding a service with requests so that others cannot use it.

The Bitcoin whitepaper, dated October 31, 2008, and published under the name Satoshi Nakamoto, describes a proof-of-work design similar to Hashcash. Bitcoin uses it so that a network of computers can agree on one chain of blocks without a central party.

How Does Proof of Work Work in Bitcoin?

Bitcoin mining is the process of creating valid blocks, and creating a valid block requires demonstrating proof of work. A miner is a participant that tries to find a valid block. A miner starts by collecting waiting transactions into a candidate block, which is a batch of transactions proposed for the chain. Next, the miner fills in the block’s header, an 80-byte summary of that block.

Bitcoin hashes the header with SHA-256 twice, a method called double SHA-256. SHA-256 is a cryptographic hash function. A hash function turns data of any size into a fixed-length fingerprint called a hash. Proof of work sets one condition: read as a number, the header’s hash must be at or below the target.

The target is a threshold that a valid block’s hash must not exceed. Network difficulty measures how much harder the current target is to hit than the easiest target the rules allow. When the target falls, network difficulty rises, and a lower target leaves fewer acceptable hashes, so the search takes longer.

To get a different hash, a miner changes the nonce, a 32-bit number stored in the header. The miner then hashes the header, compares the result with the target, and repeats. This search is often called a hash puzzle, but nothing is solved by cleverness: on average, a hash at or below the target can only be found through a corresponding amount of brute-force work, which means trying value after value.

The table below gives an example of the same search at toy scale. The rule is made up, it is not Bitcoin’s real rule, and it applies SHA-256 once instead of twice. The SHA-256 digest of the text “hello” followed by a nonce must begin with a set count of zeros. Digests are written in hexadecimal, a base-16 notation that uses the digits 0-9 and the letters a-f. A digest that begins with zeros is a small number, so each rule works as a target.

Rule

Input (text plus nonce)

SHA-256 digest

Result

Digest begins with 000

hello0

5a936ee19a0cf3c70d8cb0006111b7a52f45ec01703e0af8cdc8c6d81ac5850c

Fails (starts with 5)

Digest begins with 000

hello1

91e9240f415223982edc345532630710e94a7f52cd5f48f5ee1afc555078f0ab

Fails (starts with 9)

Digest begins with 000

hello2

87298cc2f31fba73181ea2a9e6ef10dce21ed95e98bdac9c4e1504ea16f486e4

Fails (starts with 8)

Digest begins with 000

hello10284

0006bc9ad4253c42e32b546dc17e5ea3fedaecdabef371b09906cea9387e8695

Passes on attempt 10,285

Digest begins with 0000

hello60067

0000e49eab06aa7a6b3aef7708991b91a7e01451fd67f520b832b89b18f4e7de

Passes on attempt 60,068

In this toy search, each extra zero multiplies the average number of attempts by 16. A three-zero rule takes about 4,096 attempts on average and a four-zero rule about 65,536, but a single search can take more or fewer, as the two passing rows show: 10,285 is more than 4,096, and 60,068 is fewer than 65,536. Checking any one of these results takes a single SHA-256 computation.

Bitcoin’s real search is far larger. The expected number of hashes per block is about network difficulty multiplied by 2 to the power of 32. As of October 2026, that is more than 500 sextillion hashes per block on average, and the figure changes each time network difficulty is adjusted. A sextillion is a 1 followed by 21 zeros.

Checking a block’s work is far cheaper than producing it. A node is a computer on the Bitcoin network that checks blocks against the network’s rules. To verify a block’s work, a node recomputes the double SHA-256 hash of the block header once and compares it with the target. Producing the block took that many guesses, and checking it takes one hash computation. That gap is what makes the work a proof.

Proof of Work Lets Nodes Agree on One Chain without a Central Party

Two miners can find valid blocks at nearly the same moment, which leaves the network with competing versions of the chain. Proof of work gives every node a rule for choosing: a node follows the chain with the most work behind it, which is the chain that would be hardest to recreate. Because anyone can check the work, no central party has to say which chain counts.

Each block contains the hash of the block before it, so changing a past block means redoing the work for that block and for every block after it. Proof of work is designed to make rewriting history costly, and the cost of changing a particular block grows with each block added after it.

Proof of work also bears on double spending, the attempt to spend the same BTC twice. Reversing a payment that is already in a block would mean redoing that work.

Miners Receive the Subsidy and Fees, and Network Difficulty Sets the Pace

In Bitcoin, a miner whose valid block the network accepts collects the block reward. The reward has two parts: the block subsidy, which is newly created BTC, and the transaction fees paid by the users whose transactions are in that block. As of October 2026, the subsidy for miners has been 3.125 BTC per block since the latest halving in April 2024.

The block subsidy is halved every 210,000 blocks, an event called halving that comes about every four years. Only the subsidy is halved, and the fees are not.

Every 2,016 blocks, the Bitcoin protocol adjusts network difficulty. It measures how long those blocks took, using the times recorded in their headers, and compares that with 1,209,600 seconds, which is two weeks. It then rescales the target in proportion, so that blocks keep arriving about every 10 minutes on average. A single adjustment can change network difficulty by at most a factor of 4 in either direction.

Hashrate is the total number of hashes per second that the network’s machines compute. When hashrate rises, blocks tend to arrive faster than every 10 minutes, and the next adjustment raises network difficulty. When blocks arrive too slowly, the adjustment lowers network difficulty. The adjustment measures block times only: it doesn’t count miners, and it doesn’t read the price of BTC.

Bitcoin’s 10-minute block time is an average, not a promise for each block. In a statistical model of block finding, about 63% of blocks arrive within 10 minutes of the one before.

How Does Proof of Work Differ from Proof of Stake?

Proof of work and proof of stake are both consensus mechanisms that decide which participant adds the next block to a blockchain. A blockchain is a shared record made of blocks linked in order. Proof of work ties that right to computing work: a miner hashes until a block meets the target.

Proof of stake ties that right to cryptocurrencies that a participant locks up as collateral, which means the coins stay set aside while the participant takes part in adding blocks. A participant that attempts illegitimate activity can be penalized.

The two mechanisms differ in what a participant commits: computing work in proof of work, locked coins in proof of stake.

Why Does Proof of Work Consume Electricity?

Bitcoin mining uses electricity because proof of work is real computation, and machines that compute draw power. The cost is part of the definition: a proof of work takes trial and error to produce, and in Bitcoin machines do that trial and error.

Every hash a machine computes uses some power, so more hashes in total generally means more power drawn. If more machines join, blocks tend to arrive faster, network difficulty rises at the next adjustment, and each block then takes more hashes on average.

Some policymakers and grid planners have raised concerns about how the electricity that cryptocurrency miners in the United States use affects cost, reliability, and emissions. Published estimates of Bitcoin’s electricity use differ by method and by the period they cover.

A Valid Hash Shows Effort Was Spent, Not That a Transaction Is Valid

A hash at or below the target shows that someone did the work. It shows nothing about whether the transactions inside the block are valid. A full node accepts a block only if every transaction in it is valid and not already spent, and node software such as Bitcoin Core rejects a block that claims more than the subsidy plus fees, however much work stands behind it.

Proof of work also does not make history permanent. It makes rewriting history cost work. A party that controls most of the network’s hashrate can revise recent transaction history and stop new transactions from confirming. Waiting for more confirmations, which are blocks built on top of the block that records a transaction, raises the cost of revising it without removing the possibility.

Proof of work doesn’t pay for effort either. At the protocol level, hashing that does not end in an accepted block receives no block reward, however much electricity it used.

November 27, 2025

GoMining News

You might be interested in: