What Is Bitcoin Mining? A Simple Explanation

Bitcoin mining is the proof-of-work process where specialised machines repeatedly SHA-256 a candidate block header, racing to produce a hash below the network's current target. The winner broadcasts the block, earns the 3.125 BTC subsidy plus all included transaction fees, and converts electricity into ledger security. Mining issues new BTC, orders transactions, and makes rewriting history prohibitively expensive (source: Bitcoin.org).

Mining at a glance, what actually matters in 2026:

1. Miners build candidate blocks from transactions sitting in the Bitcoin mempool and double-SHA-256 the 80-byte header until a valid hash appears

2. The block subsidy is 3.125 BTC after the April 2024 halving; the next halving cuts it to 1.5625 BTC near block 1,050,000 in April 2028 (source: Bitbo)

3. May 2026 network hashrate sits near 1,053 EH/s with difficulty around 132.47 trillion, both off the November 2025 all-time high as weaker miners exit (source: CoinWarz)

4. Modern ASICs like the Antminer S21 Hydro (335 TH/s, 16 J/TH) and Whatsminer M60S (186 TH/s, 18.5 J/TH) set the efficiency floor; everything older is under economic pressure

5. Mining is a competitive industry, not passive income, and profitability swings with electricity cost, BTC price, difficulty, and fee congestion

This article walks through mining the way a beginner most often asks about it: what miners actually do, why proof of work exists, how block rewards get paid, and why consumer hardware cannot compete in 2026. For vocabulary gaps, the Bitcoin glossary covers terms like nonce, target, and coinbase transaction before going deeper.

Why Bitcoin needs mining at all

Mining exists because Bitcoin has no central administrator. No bank, no server, no corporation decides which transactions are valid or in what order they occurred. That leaves a real design problem: how does a global network of strangers agree on one shared transaction history with no referee?

Proof of work is the answer, and mining is how proof of work gets done. The design insight is blunt. If reaching agreement costs nothing, any participant can flood the network with fraudulent histories for free. By requiring each new block to be backed by computational work that consumes real electricity and real time, the protocol makes history-rewriting genuinely expensive. An attacker who wants to reverse confirmed transactions has to outspend the honest network's cumulative work, and the deeper a transaction is buried, the more energy that would cost.

Mining therefore does two jobs at once:

• Block production. Miners assemble transactions into a canonical ordered sequence, giving every node the same view of who paid whom and when.

• Security provision. Each block burns real-world energy, which makes rewriting the recent past a quantifiable electricity bill rather than a free action.

The chain-of-hashes structure in the Bitcoin blockchain is only meaningful because producing each block requires real work. Without that cost, any participant could generate alternate histories at leisure. For the consensus mechanism on its own terms, see proof of work.

How Bitcoin mining works, step by step

Building a block template

Before any hashing starts, the miner constructs a candidate block from three pieces:

• A batch of transactions selected from the miner's own view of the mempool, almost always weighted toward higher fee rates per virtual byte

• A coinbase transaction as the very first entry, a special transaction that mints the block subsidy and sweeps all included transaction fees to an address the miner controls

• An 80-byte block header that summarizes everything above

The coinbase transaction is where new BTC enters circulation, and it is the mechanical reason any miner bothers to spin up hardware. It has no inputs, so nothing has to be spent; its output is the miner's claim to the reward. The newly minted coins from a coinbase transaction cannot be spent until the block earns 100 additional confirmations, a protocol rule that protects the network from miners double-spending their reward across short-lived chain reorganisations (source: Learn Me a Bitcoin). For context on how ordinary user transactions flow before they ever reach a miner, see how Bitcoin transactions work.

The block header and the double-SHA-256 loop

The 80-byte block header packs six fields: protocol version, previous block hash (the chain linkage), Merkle root of included transactions, timestamp, compact difficulty encoding (nBits), and a 32-bit nonce. Miners run this header through SHA-256 twice to produce a 256-bit block hash, then compare the result to the current target (source: Bitcoin Dev Docs).

SHA-256 has two properties that make the lottery work. The same input always produces the same output, so any verifier can recheck instantly. And a one-bit change to the input produces a completely different output with no predictable pattern, so miners cannot plan their way to a winning hash. The only way forward is to change a mutable field, hash again, and check whether the result fell below the target.

The target is a threshold number the protocol adjusts to keep block times near 10 minutes. A block hash is valid only if it is numerically less than the target. Because SHA-256 outputs are uniformly distributed, lower targets are statistically harder to hit, and that is exactly how difficulty scales (source: Bitcoin Wiki).

The nonce, the extraNonce, and why it is pure brute force

The primary mutable field is the 32-bit nonce. Miners increment it, hash the header, check the result, and repeat. At modern hashrates the 4.3-billion-wide nonce space exhausts in a fraction of a second, so miners also rotate a counter embedded in the coinbase transaction called the extraNonce, which shifts the Merkle root and produces a fresh stream of header candidates to hash. The search has no memory and no shortcut. It is a brute-force lottery where every draw is independent.

When a miner finally produces a header hash below the target, they broadcast the block immediately. Every full node independently verifies it within milliseconds: header hash below target, correct coinbase subsidy, all transactions valid under consensus rules. If any check fails, honest nodes reject the block regardless of who mined it. This is why mining secures ordering but does not grant rule-making power, a distinction developed in nodes vs miners vs wallets.

Hashrate, difficulty, and the 10-minute rhythm

What hashrate measures

Hashrate is the number of SHA-256 attempts executed per second. A single modern ASIC runs at hundreds of terahashes per second (TH/s); the whole network in May 2026 runs near 1,053 EH/s, or roughly 1.1 quintillion attempts every second (source: CoinWarz). The figure moved significantly through late 2025 and early 2026, hitting a November 2025 all-time high near 155 trillion difficulty, then falling through the first quarter of 2026 as weaker miners shut down under post-halving pressure and rising electricity costs (source: The Block).

Higher network hashrate does not guarantee any individual miner wins more blocks. A miner's expected share of blocks equals their share of total hashrate. A 1% miner wins roughly 1% of blocks over a long enough horizon, but on any given day a 0.01% miner could get lucky and win twice in a row. Every hash is an independent Bernoulli trial with identical probability. This variance is the entire reason mining pools exist.

Difficulty adjustment

Every 2,016 blocks, approximately every two weeks, the protocol checks how long those 2,016 blocks actually took. If they arrive faster than two weeks, the target tightens (difficulty goes up). If they arrive slower, the target loosens (difficulty goes down). The adjustment holds a block interval near 10 minutes regardless of how much hashrate enters or exits the market.

The first quarter of 2026 demonstrated this mechanism plainly. Difficulty dropped 11.16% in February and another 7.76% in late March, the second-largest adjustments of the year, as miners capitulated and hashrate followed them offline The Block. The 10-minute target held through both adjustments. What difficulty does NOT do is guarantee profitability or compensate for rising electricity costs; it only governs block timing. For the retarget arithmetic in depth, see Bitcoin difficulty adjustment explained.

What miners earn: Subsidy, fees, and the halving schedule

The coinbase transaction pays winners two things. The block subsidy is newly created BTC on a fixed declining schedule. Transaction fees are the sum of fees paid by every transaction the miner chose to include. Together these make the block reward.

The 210,000-block halving cycle has now executed four times:

The April 2028 halving will cut daily BTC issuance from roughly 450 to roughly 225 (source: Bitbo). The hard cap is 21 million total BTC, enforced by consensus rules; subsidy reaches effectively zero around the year 2140. For the supply math in detail, see Bitcoin's 21 million supply cap.

Fee income matters more than it used to. In early 2024, Ordinals inscription activity pushed transaction fees to historically high shares of block reward. Fee revenue is still volatile and mempool-dependent, but it is also the only reward stream that grows if subsidy shrinks. As each halving compresses subsidy, miners become structurally more reliant on fees, which puts more pressure on congestion-driven fee markets during periods of high usage. For the pattern of when fees spike, see why Bitcoin fees get high.

Mining economics: J/TH, electricity, and why 2026 is painful

Mining profitability is not stable. It depends on several variables that all move independently, and each one can flip a comfortable operation into loss territory. Coming out of the 2024 halving and into 2026, the pressure is showing in the data. Mining companies reportedly faced cost-per-coin figures materially above prevailing BTC price during parts of the first quarter (source: CoinDesk), and the successive difficulty downticks reflect weaker operators switching off.

No single profit figure makes sense without fixing all seven variables, and any mining calculator that omits depreciation, downtime, or difficulty trend is showing an incomplete picture. A spreadsheet at constant difficulty typically flatters returns; real operators model difficulty growth and fee volatility before signing hardware contracts.

From BloFin's exchange-operator perspective, we watch miner flows alongside exchange deposit patterns, and the behavioural signal during the 2026 difficulty drops was consistent with miner capitulation: hashpower retreating as marginal operators switched off rather than continuing to mine at a loss. That flow dynamic is why mining is a business requiring active management, not a passive yield.

For whether a specific setup can work, use is Bitcoin mining profitable.

For the environmental and policy angles behind the electricity demand, see Bitcoin and energy alongside the energy debate.

YMYL note: Bitcoin mining is capital-intensive and profitability is uncertain. Before committing to hardware or electricity contracts, independently model electricity cost, hardware efficiency, hardware purchase price, projected difficulty growth, and BTC price scenarios including significant declines. This page is education, not investment advice.

Hardware generations: CPU, GPU, FPGA, ASIC

Mining hardware has moved through four generations, each obsoleting the last on efficiency per watt.

• CPU (2009-2010). Satoshi Nakamoto and early adopters mined with regular laptops and desktops. Network hashrate was negligible, difficulty sat near its minimum, and the subsidy was 50 BTC. Any PC could win blocks.

• GPU (2010-2012). Graphics cards have hundreds of parallel cores designed for bulk numerical operations; they hash far more efficiently than CPUs. GPU mining rendered CPU mining uncompetitive within months.

• FPGA (2011-2013). Field-Programmable Gate Arrays were a brief middle step, more efficient than GPUs but harder to deploy.

• ASIC (2013-present). Application-Specific Integrated Circuits are silicon designed from scratch to do exactly one thing, compute SHA-256 as fast per watt as physics allows. The first ASICs pushed GPU and FPGA mining into economic irrelevance within a year.

Modern 2026 ASICs set the current efficiency floor. The Bitmain Antminer S21 Hydro uses immersion cooling to reach 335 TH/s at roughly 16 J/TH (source: Bitmain). The air-cooled Antminer S21 hits 200 TH/s at 17.5 J/TH. The Whatsminer M60S runs at 186 TH/s and 18.5 J/TH (source: D Central). A GPU attempting SHA-256 mining today consumes thousands of times more electricity per terahash than any of those, which is why consumer hardware mining is not economically relevant for Bitcoin.

A laptop, phone, or gaming PC cannot mine Bitcoin competitively. The hashrate contribution is so small that expected daily earnings are a fraction of a cent, far less than the electricity consumed, and the sustained thermal load can damage hardware not built for continuous maximum load. Home ASIC mining is a different question, addressed separately in Bitcoin home mining. For the full ASIC generation history and specs, see Bitcoin mining ASICs.

Solo mining vs mining pools: The variance problem

Why solo mining is a lottery most operators cannot afford

With a total network hashrate near 1,053 EH/s, a single modern ASIC at 200 TH/s represents roughly 0.00002% of the network. The expected number of blocks such a machine wins in a year is a fraction, and actual timing is random. A solo miner could wait months or years between wins, carrying electricity costs the entire time. The expected value exists on paper; few operators can sustain the variance.

What pools actually do

A mining pool aggregates hashrate from many participants and submits share proofs to a central pool server that tracks each member's contribution. When the pool wins a block, the reward is distributed among participants according to the pool's payout model. This converts an unpredictable lottery into a more predictable stream of smaller payouts without changing the expected earnings per unit of hashrate.

Three payout models dominate:

• PPS (Pay-Per-Share). The pool pays a fixed amount for every valid share the miner submits, regardless of whether the pool wins. Predictable, but basic PPS excludes transaction fees from miner payouts.

• FPPS (Full-Pay-Per-Share). The pool extends PPS to include an average of expected transaction fees in every share payout. FPPS is the dominant model in 2026 because it gives miners the predictability of PPS plus fair access to fee revenue (source: Minebest).

• PPLNS (Pay-Per-Last-N-Shares). The pool pays only when it wins a block and distributes based on the last N shares. Miner income is more volatile (pool luck matters), but pool fees are typically lower. PPLNS suits large-scale operators with capital to tolerate variance.

The full mechanics of each scheme are covered in PPS vs PPLNS. The broader pool landscape, including how to choose a pool, is in what are Bitcoin mining pools.

What pools do NOT change

Joining a pool does not make mining easier at the network level. Difficulty adjusts based on total hashrate, not on how that hashrate is organised. The pool only smooths out a miner's income stream. The pool operator usually assembles the block template and controls the coinbase transaction, which is why pool concentration raises decentralisation concerns. In 2026 those concerns are visible: Foundry USA and AntPool together account for roughly 55-60% of network hashrate, and the top four pools control the large majority of blocks (source: Hashrate Index). For the centralization debate, see is Bitcoin centralized.

2026 landscape: Geography, energy, and concentration

Mining looked different in 2019 than in 2026 because two large shifts played out in between. First, the June 2021 China ban triggered the fastest voluntary geographic reshuffle in Bitcoin's history, as roughly half the network's hashrate migrated to the United States, Kazakhstan, Russia, and other jurisdictions over several months. Second, institutional capital, public-company mining operations, and large-scale facility co-location with flexible power sources have consolidated the industry.

The 2026 picture from Cambridge Centre for Alternative Finance and Hashrate Index data:

• United States leads global hashrate at roughly 35-40% share, with Texas hosting a large fraction due to the ERCOT deregulated power market, Georgia and Kentucky maintaining meaningful shares, and New York holding a smaller slice reflecting BitLicense friction (source: Hashrate Index).

• Energy mix. The Cambridge 2025 update put the sustainable share of Bitcoin's electricity at approximately 52.4%, made up of roughly 42.6% renewables (hydro, wind, solar) plus 9.8% nuclear, with fossil fuels (mainly natural gas) at the remainder (source: Cambridge CJBS). The Bitcoin Mining Council tends to report higher sustainable percentages using a different methodology, so estimates vary by source. Gas flaring, stranded hydro, and grid-balancing curtailment arrangements are the new business frontier.

• Pool concentration. Foundry USA at roughly 30% and AntPool at roughly 20-25% dominate; ViaBTC and F2Pool fill out the top four (source: Cryptoslate).

• AI competition. Rising energy prices, compounded by AI data-centre demand bidding for the same grid capacity, have compressed miner margins through 2025 and 2026. Some public miners have announced AI compute pivots to diversify revenue.

From Blofin's deposit-engineering perspective, miner withdrawal patterns are a meaningful flow signal. During the Q1 2026 difficulty drops, we observed the usual post-halving exchange-inflow pattern from miner wallets, consistent with operators converting reward into fiat to cover electricity invoices rather than holding through the drawdown. The mining business cycle and the exchange-deposit cycle are tightly linked in ways that pure price charts miss. For the grid-level story, see Bitcoin mining and power grids.

What mining secures and what it does not

The security guarantee

Mining builds a chain where each accepted block requires real energy. Rewriting history means redoing the proof of work for a target block and every block stacked on top of it, producing a chain with more cumulative work than the honest chain, all while the honest network keeps extending its own tip. At 1,053 EH/s, that attack is priced out for any single actor in recent history, and the deeper a transaction is buried, the more expensive rewriting it becomes.

What a 51% attacker could and could not do

A majority-hashrate attacker could, in theory:

• Privately mine a longer chain and release it, reversing recent transactions they were party to (double-spends)

• Exclude specific transactions from blocks they mine (censorship)

A majority attacker cannot:

• Mint BTC beyond the protocol schedule (the consensus rule lives in nodes, not miners)

• Steal funds from addresses whose keys they do not control

• Rewrite deeply confirmed history without an impractical energy bill

• Change Bitcoin's consensus rules

The boundary matters. Mining orders transactions and makes recent history costly to rewrite. Full nodes, not miners, enforce the rule set, and they reject invalid blocks regardless of hashrate behind them. For the attack class in depth, see what is a 51% attack.

Mining vs buying Bitcoin

For most readers, the question under "should I mine" is really "should I mine, or should I just buy?" The honest 2026 answer is: unless you have a concrete structural advantage (cheap industrial electricity, cooling infrastructure, capital to buy at wholesale, operational expertise), buying on an exchange is the cleaner path to BTC exposure. Mining profitability requires outperforming an industry where public companies run hundreds of thousands of ASICs with sub-3-cent power and treasury-scale hardware purchasing.

Where mining still makes sense in 2026:

• Strategic energy positioning. Access to below-market power, gas-flaring opportunities, stranded hydro, or grid-balancing curtailment agreements.

• Home hobbyist mining with realistic expectations. A single S21 in a garage is a hobby expense that produces a small BTC trickle, not a wealth-building plan, which is a separate conversation from the industrial economics outlined above.

• Operational edge. Existing facilities, engineering teams, or financing structures that a competitor cannot replicate cheaply.

Everything else rewards the buyer over the miner on a risk-adjusted basis in 2026. A miner carries capital depreciation, difficulty growth, BTC price exposure, electricity rate exposure, hardware obsolescence, and execution risk; a buyer carries only price exposure.

Frequently asked questions

What is Bitcoin mining in simple terms?

Bitcoin mining is the process where specialised computers compete to group recent transactions into a block and add it to Bitcoin's transaction history. They do this by running the SHA-256 hash function on a block header billions of times per second until they produce an output below the network's current difficulty target. The first miner to find a valid result broadcasts the block and earns the block reward, which is newly issued BTC plus all the transaction fees in that block. Mining is Bitcoin's proof-of-work mechanism, and its energy cost is what secures transaction history.

How does Bitcoin mining actually work step by step?

A miner assembles a candidate block from pending mempool transactions, adds a coinbase transaction that claims the subsidy plus fees, and constructs an 80-byte block header. The miner hashes that header using double SHA-256, changing a 32-bit nonce (and the coinbase extraNonce when the nonce space runs out) between attempts until the resulting hash falls below the current target. When that happens, the block is broadcast, full nodes verify it independently, and it is appended to the chain. Difficulty adjusts every 2,016 blocks to keep the average block interval near 10 minutes regardless of how much hashrate joins or leaves.

What is the current Bitcoin block reward in 2026?

The block reward in 2026 has two parts. The block subsidy, after the April 2024 halving, is 3.125 BTC per block. Transaction fees are the sum of fees paid by every transaction the miner included, which varies with mempool congestion. Combined, these are paid to the miner through the coinbase transaction in each block. Subsidy halves every 210,000 blocks; the next halving is projected for around 12 April 2028 at block 1,050,000 and will reduce subsidy to 1.5625 BTC, cutting daily BTC issuance from roughly 450 to roughly 225.

Why does Bitcoin mining use so much energy?

Energy consumption is not a side effect of mining; it is the mechanism. Proof of work is deliberately designed so that producing a valid block requires burning real electricity that cannot be faked or simulated cheaply. That cost is what makes Bitcoin's security guarantee meaningful, because an attacker wanting to rewrite history has to redo the same energy-intensive work faster than the honest network continues extending it. Higher network hashrate means more total energy used but also a higher cost of attack. The debate about renewable share, grid impact, and opportunity cost is separate from the protocol reason the energy gets burned.

Is Bitcoin mining still profitable in 2026?

Profitability depends on variables that change continuously: BTC price, network difficulty, electricity cost, hardware efficiency, hardware purchase price, and transaction fee levels. Operators with sub-$0.05/kWh power and modern ASICs like the S21 Hydro can still operate profitably; operators with older hardware or higher power costs have faced losses through parts of Q1 2026. The successive difficulty drops in February and March 2026 reflected weaker miners switching off under that pressure. No single profit figure applies without specific assumptions about every input, and any mining calculator that omits depreciation, downtime, and difficulty trend flatters returns.

Can I mine Bitcoin on a laptop, phone, or gaming PC in 2026?

No, not competitively. Modern Bitcoin mining requires ASICs, purpose-built SHA-256 chips that outperform any general-purpose CPU or GPU by orders of magnitude in hashes per watt. A laptop running mining software would spend more on electricity than it would ever earn in BTC, and sustained thermal load can damage hardware not built for continuous full load. Home ASIC mining is technically possible with dedicated electrical circuits and cooling, but whether it is economical depends on local electricity rates, hardware price, and network difficulty. At typical residential electricity prices in 2026, home ASIC mining is rarely profitable after depreciation.

Why do miners join pools instead of mining solo?

A modern ASIC at 200 TH/s is about 0.00002% of the 2026 network, so solo mining is a lottery where the operator could wait months between wins while paying electricity throughout. Pools aggregate hashrate and pay out proportionally when the pool wins a block, trading occasional large solo wins for a steady stream of smaller payouts. Expected earnings per unit of hashrate are the same minus the pool fee, but variance is far lower. The tradeoff is pool operator control over templates, which is why Foundry plus AntPool at 50%+ hashrate is a live decentralisation concern.

Researched and written by the BloFin Academy editorial team with AI-assisted drafting. All facts independently verified.

Disclaimer: This content is for educational purposes only and does not constitute financial, investment, legal, or tax advice. Crypto assets are highly volatile and carry significant risk of loss. Always verify local regulations and consult a qualified professional before making financial decisions.