How Does Bitcoin Mining Work?

By Blockstream Engineering · Last updated June 2026

What Is Bitcoin Mining?

Bitcoin mining is the backbone of the Bitcoin network. Every Bitcoin transaction that has ever been confirmed was processed by a miner, and every block in the blockchain was proposed by a miner. The roughly 20 million bitcoin in circulation were all created through mining.

At a high level, mining accomplishes two things simultaneously:

1. Transaction processing: Miners select unconfirmed transactions from the mempool and package them into blocks. Once a block is accepted by the network, the transactions inside it become part of Bitcoin's permanent ledger.
2. New bitcoin issuance: Each valid block includes a special transaction called the coinbase transaction, which creates new bitcoin and assigns them to the miner. This is the only mechanism by which new bitcoin enters circulation.

The word "mining" is an analogy. Gold miners expend physical labor and energy to extract gold from the earth. Bitcoin miners expend computational power and electrical energy to produce new bitcoin and secure the network. The analogy holds because in both cases, the resource cost is real and the output is scarce.

Unlike gold mining, Bitcoin mining simultaneously processes and secures a global payment network.

Why Does Bitcoin Need Mining?

Mining is what allows a decentralized system to agree on a single, authoritative transaction history without trusting any individual participant. Bitcoin has no central authority: no bank approves transactions and no company runs the servers. The network operates across tens of thousands of nodes spread across every continent.

Transaction Confirmation

When you send bitcoin, your transaction is broadcast to the network and sits in the mempool (memory pool), a waiting area of unconfirmed transactions. It stays there until a miner includes it in a block. Without miners, no transactions get confirmed. The bitcoin you sent would sit in limbo permanently.

Miners prioritize transactions by fee rate (measured in satoshis per virtual byte). Higher-fee transactions get included faster because miners are economically incentivized to maximize their revenue per block. This creates a market-based system for block space: when demand is high, fees rise; when demand is low, fees drop.

Network Security

Mining is also Bitcoin's security system. To alter a confirmed transaction, an attacker would need to redo the proof-of-work for the block containing that transaction and every subsequent block, faster than the rest of the network continues building new blocks. With the global hashrate exceeding 800 exahashes per second (EH/s) as of Q1 2026 (per mempool.space / Coin Metrics), this requires an investment of billions of dollars in specialized hardware and the ongoing energy to power it.

The deeper a transaction is buried under subsequent blocks, the more work an attacker would need to redo. This is why merchants and exchanges often wait for multiple confirmations (typically six, about one hour) before considering a large payment final.

How Miners Build a Block

Before a miner can start hashing, it needs to assemble a candidate block: a proposed block that follows all of Bitcoin's consensus rules and, if a valid hash is found, will be submitted to the network.

Selecting Transactions From the Mempool

The miner's node maintains a copy of the mempool. From this pool of unconfirmed transactions, the miner selects which ones to include. The selection criteria are straightforward:

• Validity: Each transaction must be properly formatted, have valid signatures, and spend only unspent outputs (no double-spends).
• Fee priority: Transactions paying higher fee rates (sats/vB) are selected first, since the miner keeps all transaction fees.
• Block weight limit: A block can contain up to 4 million weight units (roughly 1 MB of non-witness data, or up to about 4 MB including witness data under SegWit). The miner packs in as many high-fee transactions as the limit allows.

Constructing the Block Header

With transactions selected, the miner builds the block header. This 80-byte data structure is the actual input to the mining computation. It contains six fields:

The Merkle Root

The merkle root is a cryptographic summary of every transaction in the block. It is computed by hashing each transaction ID with SHA-256d, then pairing and hashing those hashes together, repeating the process upward until a single 32-byte root remains. The merkle tree construction is implemented in ComputeMerkleRoot() in src/consensus/merkle.cpp. This structure means that changing even one transaction in the block would produce a completely different merkle root, which would invalidate the block header's hash.

The merkle tree also enables Simplified Payment Verification (SPV), where a lightweight client can prove a transaction is included in a block by checking a short merkle proof rather than downloading the entire block.

The Proof-of-Work Puzzle

With the candidate block assembled, the miner begins the core computation: finding a nonce value that, when included in the block header and hashed, produces a result below the network's current difficulty target.

SHA-256: The Hash Function

Bitcoin uses SHA-256 (Secure Hash Algorithm, 256-bit), applied twice: SHA-256(SHA-256(block_header)), commonly written as SHA-256d. The double-hashing is used throughout the protocol for block headers, transaction IDs, and merkle trees. This function takes any input and produces a fixed 256-bit (32-byte) output that appears random. Key properties:

• Deterministic: The same input always produces the same output.
• Avalanche effect: Changing a single bit of input completely changes the output. There is no way to predict how the output will change.
• One-way: Given a hash output, there is no way to compute the input. The only strategy is brute force: guess an input, hash it, and check.
• Collision resistant: It is computationally infeasible to find two different inputs that produce the same hash.

What Miners Are Actually Computing

The miner takes its 80-byte block header, applies SHA-256d, and checks whether the resulting 256-bit hash is numerically less than the difficulty target. If not, the miner increments the nonce by one and hashes again. And again. And again.

A 256-bit hash can be thought of as a number between 0 and 2²⁵⁶ (a number with 77 digits). The difficulty target specifies a threshold: the hash must start with a certain number of leading zeros. The more leading zeros required, the smaller the valid range, and the harder it is to find a qualifying hash.

At current difficulty levels (early 2026), only about 1 in every 10²³ hash attempts will produce a valid result. That is roughly the probability of randomly selecting a specific atom from a teaspoon of water.

Beyond the 4-Byte Nonce

The nonce field is only four bytes, giving 2³² (about 4.3 billion) possible values. Modern ASICs exhaust the entire nonce space in well under a second. To keep searching, miners modify other parts of the block to create new header permutations:

• ExtraNonce: A field in the coinbase transaction that the miner can freely modify. Changing the coinbase transaction changes the merkle root, which changes the block header, creating an entirely new set of 4.3 billion nonce values to try.
• Timestamp: Miners can adjust the timestamp within an allowed range, producing new header permutations.
• Transaction selection: Adding, removing, or reordering transactions changes the merkle root.

Trivial to Verify, Impossible to Shortcut

The fundamental asymmetry of proof-of-work is that verification is instant. When a miner broadcasts a new block, every node on the network hashes the block header once, checks that the result is below the target, and knows immediately whether the block is valid. The miner spent trillions of attempts to find that hash; the network needs exactly one computation to verify it.

There is no shortcut. No mathematical formula can predict which nonce will produce a valid hash. The only path is raw computation: guess, hash, check, repeat. This is what makes proof-of-work a fair competition. A miner with 10% of the global hashrate will, on average, win 10% of the blocks, regardless of any other advantage.

Block Validation

Finding a valid hash is only half the process. The miner broadcasts its block to the network, and every node independently verifies it before adding it to their copy of the blockchain.

What Nodes Check

Full nodes perform a comprehensive validation of every proposed block. The checks include:

• Proof-of-work: The block header hash is below the current difficulty target.
• Block structure: The block follows the correct format and does not exceed the weight limit.
• Previous block hash: The block correctly references the hash of the most recent valid block.
• Timestamp: The timestamp is within acceptable bounds (not too far in the future, not before the median of the last 11 blocks).
• Coinbase transaction: The block reward does not exceed the allowed amount (subsidy + fees). The subsidy is computed by GetBlockSubsidy() in src/validation.cpp, which right-shifts the initial 50 BTC reward by the number of completed halvings. The coinbase maturity rule requires that coinbase outputs cannot be spent for 100 blocks, enforced by the COINBASE_MATURITY constant in src/consensus/consensus.h.
• Transaction validity: Every transaction in the block has valid signatures, spends only existing unspent outputs, and follows all script rules.
• No double-spends: No output is spent more than once within the block or across the blockchain.

If any check fails, the node rejects the block entirely. There is no partial acceptance. A block with one invalid transaction is as invalid as a block with a forged proof-of-work. This is Bitcoin's consensus enforcement: every node acts as an independent auditor, and no amount of hashrate can force an invalid block onto the network.

Propagation and Convergence

Valid blocks propagate across the network in seconds. Nodes relay blocks to their peers, who validate and relay further. Compact block relay (BIP 152) speeds this process by transmitting only a summary of the block's transactions, since most nodes already have the transactions in their own mempools.

Occasionally, two miners find valid blocks at nearly the same time, creating a temporary fork. The network resolves this naturally: whichever chain gets the next block added to it first becomes the longest chain, and nodes switch to it. The "orphaned" block's transactions return to the mempool to be included in a future block. No transactions are lost.

Even if a miner loses internet connectivity, alternative methods exist for receiving block data. Blockstream Satellite, a satellite broadcast service Blockstream previously operated (transferred to an independent mining entity in April 2025), broadcasted the Bitcoin blockchain from space, providing a backup data source for miners and nodes that needed to stay synchronized without relying solely on terrestrial internet connections.

The Difficulty Adjustment

Bitcoin targets an average block time of ten minutes. The total computational power (hashrate) directed at mining changes constantly as miners join, leave, upgrade equipment, or go offline. The difficulty adjustment is the protocol mechanism that maintains the ten-minute target regardless of these fluctuations.

How It Works

Every 2,016 blocks (approximately two weeks at ten minutes per block), every node on the network performs the same calculation:

1. Measure how long the last 2,016 blocks actually took to mine.
2. Compare that duration to the expected duration (2,016 blocks × 10 minutes = 20,160 minutes, or exactly two weeks).
3. Adjust the difficulty target proportionally. If blocks came faster than expected, the target decreases (making it harder). If blocks came slower, the target increases (making it easier).

The adjustment is capped at a factor of four in either direction per period, preventing extreme swings. This cap is enforced in CalculateNextWorkRequired() in src/pow.cpp, which clamps the actual timespan to the range [20,160 minutes / 4, 20,160 minutes × 4] before computing the new target. In practice, adjustments are usually small (single-digit percentages) because hashrate changes gradually between periods.

Under the Hood: The Difficulty Target Calculation

The difficulty target is a 256-bit number stored in the block header as a compact 4-byte representation called nBits. A block hash must be numerically less than this target to be valid. The relationship between the human-readable "difficulty" number and the target is inverse: higher difficulty means a lower target, meaning fewer valid hashes exist in the 2²⁵⁶ output space.

The target for the genesis block (difficulty 1) was 0x00000000FFFF0000000000000000000000000000000000000000000000000000. The current difficulty can be derived from the current target: difficulty = genesis_target / current_target. At difficulty above 130 trillion (as of mid-2026), the valid hash space is roughly 1 in 10²³ of the total 2²⁵⁶ output space.

The retargeting formula itself is straightforward: new_target = old_target × (actual_time / expected_time), where expected_time = 2,016 × 600 seconds. The target block time of 600 seconds is set by nPowTargetSpacing in src/kernel/chainparams.cpp, and the retarget interval of 2,016 blocks is defined by nPowTargetTimespan / nPowTargetSpacing. Because every node runs the same calculation on the same block timestamps, all nodes independently converge on the same difficulty for the next period.

Why This Matters

The difficulty adjustment creates a self-regulating system:

• Predictable issuance: Regardless of whether the network has 1 TH/s or 800 EH/s of hashrate, new blocks (and new bitcoin) arrive approximately every ten minutes. The supply schedule is immune to changes in mining capacity.
• Miner equilibrium: If the Bitcoin price rises, mining becomes more profitable, attracting more hashrate. Difficulty increases to match. If the price falls, less-efficient miners shut down, hashrate drops, and difficulty decreases, making mining viable again for the remaining miners. This creates a dynamic equilibrium.
• Attack resistance: An attacker cannot simply throw more hashrate at the network to produce blocks faster in the long run. Difficulty adjusts to absorb the additional power.

Mining Incentives

Miners expend real resources (hardware, electricity, facilities, maintenance) to mine blocks. The incentive to do so comes from two sources: the block subsidy and transaction fees.

The Block Subsidy

Each new block includes a coinbase transaction that creates new bitcoin. This block subsidy started at 50 bitcoin per block when Bitcoin launched in January 2009 and halves approximately every four years (every 210,000 blocks):

After the April 2024 halving, miners earn 3.125 bitcoin per block. This halving schedule is hardcoded into the protocol and will continue until approximately the year 2140, when the subsidy drops below 1 satoshi and all 21 million bitcoin have been issued.

Transaction Fees

The second revenue source is transaction fees. Every Bitcoin transaction includes a fee (the difference between the total input value and the total output value), and the miner who includes that transaction in a block collects the fee.

As the block subsidy decreases over time, transaction fees will become an increasingly important part of miner revenue. This transition is fundamental to Bitcoin's long-term security model: once the subsidy reaches zero, the network will be secured entirely by transaction fees.

In 2024, transaction fees typically accounted for 2-10% of total block revenue, though fee spikes during high-demand periods have pushed fees past 50% of block revenue in individual blocks.

Why Incentives Keep Miners Honest

A miner who produces an invalid block (claiming more bitcoin than allowed, including invalid transactions, or violating any consensus rule) will have that block rejected by every node on the network. The miner wasted the energy required to find the proof-of-work and earned nothing. Honest mining is always more profitable than dishonest mining, because dishonest blocks are rejected.

Miners do not have to be trusted, because the protocol makes honest behavior the most profitable strategy. A miner who follows the rules earns the block reward; a miner who breaks them wastes energy on a block the network rejects.

The Economics of Mining

Bitcoin mining is a capital-intensive business with thin margins. Profitability depends on the interplay of several variables, and miners must manage all of them simultaneously.

Revenue

A miner's gross revenue comes from two sources:

• Block subsidy: Currently 3.125 bitcoin per block, computed by right-shifting the initial 50 BTC reward by the number of completed 210,000-block halving intervals. After the April 2024 halving (block 840,000), the subsidy dropped from 6.25 to 3.125 bitcoin, cutting miner revenue from new issuance in half overnight. The next halving is expected in 2028.
• Transaction fees: Variable, driven by demand for block space. During periods of high network activity, fees can spike significantly. In normal conditions, fees represent 2-10% of total block revenue. During exceptional demand events (like Ordinals inscription surges in 2023-2024), fees have temporarily exceeded the block subsidy.

Costs

The major cost categories for a mining operation:

• Electricity: The dominant ongoing expense, typically 60-80% of total operating costs. Electricity prices vary enormously by geography, from under $0.03/kWh in parts of the Middle East, Africa, and Latin America to over $0.15/kWh in most of Western Europe. The difference between $0.03 and $0.08 per kilowatt-hour can determine whether an operation is profitable or hemorrhaging cash.
• Hardware: A modern ASIC miner costs $2,000 to $8,000 depending on the model and efficiency rating. Machines are typically amortized over two to four years before efficiency degradation and rising difficulty make them uncompetitive.
• Cooling and infrastructure: ASIC miners generate substantial heat. Facilities require industrial cooling systems (air cooling, immersion cooling, or hydro cooling), which add both capital and operating expenses. Building or leasing a facility with adequate power capacity, ventilation, and physical security adds further cost.
• Maintenance: Hardware failures are routine at scale. Fans fail, hash boards degrade, and power supplies burn out. Large operations employ full-time technicians and maintain spare parts inventories.

The Profitability Equation

Three external variables drive mining profitability beyond a miner's direct control:

1. Bitcoin price: Revenue is denominated in bitcoin but costs are denominated in local currency. A rising bitcoin price increases revenue in fiat terms without changing costs. A falling price does the opposite.
2. Network difficulty: The Bitcoin protocol adjusts mining difficulty every 2,016 blocks (approximately two weeks) to maintain an average block time of ten minutes. When more hashrate joins the network, difficulty rises, meaning each unit of hashrate earns a smaller share of the total block reward. When miners shut down (often due to price drops or rising costs), difficulty falls and surviving miners earn more per unit of hashrate.
3. Electricity cost: Since electricity represents the majority of operating expenses, even small differences in electricity rates compound into large profitability differences over time. This is why miners are relentlessly drawn to the cheapest energy sources available, wherever they exist geographically.

The difficulty adjustment creates a self-correcting economic cycle. When bitcoin's price rises sharply, mining becomes highly profitable, attracting new hashrate. That new hashrate increases difficulty, compressing margins back toward equilibrium. When price drops, the least efficient miners (those with high electricity costs or older hardware) shut down first. Difficulty then adjusts downward, restoring profitability for the survivors. This feedback loop ensures that Bitcoin mining gravitates toward an equilibrium where the marginal miner is barely profitable.

Mining Hardware Evolution

The history of Bitcoin mining hardware is a story of relentless specialization, driven by the economic incentive to compute SHA-256 hashes as efficiently as possible.

CPUs (2009-2010)

Satoshi Nakamoto mined the first Bitcoin blocks using a standard CPU (central processing unit). In the earliest days, a single desktop computer could mine blocks. CPUs are general-purpose processors designed for a wide variety of tasks, making them inherently inefficient at the narrow, repetitive computation that mining requires.

GPUs (2010-2012)

Graphics processing units (GPUs) contain thousands of small cores optimized for parallel computation. Since SHA-256 hashing is highly parallelizable, GPUs offered roughly 10-100x the hashrate of CPUs at better energy efficiency. GPU mining marked the first significant leap in mining specialization.

FPGAs (2011-2013)

Field-programmable gate arrays (FPGAs) are integrated circuits that can be configured after manufacturing for specific tasks. Miners programmed FPGAs to perform SHA-256 hashing, gaining moderate efficiency improvements over GPUs. FPGAs served as a bridge between general-purpose hardware and fully custom silicon.

ASICs (2013-Present)

Application-specific integrated circuits (ASICs) are chips designed from the ground up to do exactly one thing: compute SHA-256 hashes. They cannot run other software. They cannot be repurposed. This extreme specialization makes them orders of magnitude more efficient than any general-purpose hardware.

Modern Bitcoin ASICs (as of early 2026) operate in the range of 150-350+ terahashes per second (TH/s) per unit, with energy efficiency around 15-25 joules per terahash (J/TH). For comparison, a top-end GPU might achieve 1-2 GH/s at hundreds of watts. A single ASIC does what tens of thousands of GPUs cannot match.

The ASIC era ended the possibility of mining bitcoin profitably with consumer hardware. Bitcoin mining today is an industrial-scale operation requiring significant capital investment in specialized equipment, facilities, and energy contracts. The industry has matured from hobbyist origins to dedicated infrastructure companies. Blockstream operated large-scale mining facilities leveraging renewable and stranded energy sources before its mining division became an independent entity in April 2025.

Under the Hood: How ASICs Compute SHA-256d

A Bitcoin ASIC contains hundreds or thousands of identical SHA-256 "cores" etched into silicon, each executing the algorithm's 64 rounds of bitwise operations (Ch, Maj, Σ0, Σ1) in dedicated logic gates rather than loading instructions from memory. Because the block header is exactly 80 bytes, it fits into two SHA-256 compression rounds (the first 64-byte chunk and a padded second chunk). The double-hash requires running the 256-bit output of the first SHA-256 through a third compression round. Each core on the chip runs this three-stage pipeline independently, testing a different nonce value per clock cycle.

Modern ASICs are manufactured on advanced process nodes (5nm and below), allowing billions of transistors on a single die. The chip designer's primary optimization target is joules per hash: minimizing the energy each logic gate consumes per switching event. This is why efficiency (J/TH) improves with each generation even when raw hashrate gains are modest. A 5nm chip performing the same SHA-256 operation as a 16nm chip does so with a fraction of the energy per gate transition.

Understanding ASIC Specifications

Three metrics define an ASIC miner's capability. The most important is hashrate, the rate at which the machine computes SHA-256 hashes:

Efficiency (J/TH) is the single most important specification for evaluating mining hardware. Two machines with identical hashrates can have wildly different electricity costs. A miner rated at 15 J/TH will consume roughly 40% less electricity than one rated at 25 J/TH while producing the same hashrate. Over months and years of continuous operation, that difference determines whether the machine is profitable.

Major ASIC Manufacturers

ASIC manufacturing is concentrated among a few companies:

New entrants continue to emerge. Companies like Auradine (based in the US) and Bitdeer (designing their own SEAL chips) are working to diversify the supply chain. The concentration of ASIC manufacturing in a small number of companies is one of the ongoing centralization pressures in the mining ecosystem.

Solo Mining vs. Pool Mining

The economics of mining present an individual miner with a fundamental question: mine alone, or join a pool?

Solo Mining

A solo miner keeps the entire block reward (subsidy + fees) when they find a valid block. The problem is probability. With the global hashrate above 800 EH/s, a miner running a single modern ASIC at 300 TH/s controls roughly 0.0000375% of the network's total hashrate. The expected time between solo blocks is measured in decades, and the miner pays electricity costs every hour regardless.

Pool Mining

Mining pools aggregate hashrate from many miners, reducing variance. When the pool finds a block, the reward is distributed among participants based on their contributed work.

How Pools Track Contributions

Pools assign work to each miner at a lower difficulty than the network target. These easier "shares" serve as statistical proof of how much hashing each miner is performing. A miner who submits twice as many shares is doing twice the work and earns twice the payout when the pool finds a block.

Pool Payout Methods

Different pools use different methods to distribute rewards:

Most miners choose FPPS pools because the predictable revenue stream simplifies operational planning. PPLNS pools can be slightly more profitable over long periods (because the pool operator takes a smaller cut), but the payout variance makes them harder to budget around.

Why Most Miners Use Pools

Solo payouts arrive too rarely for most miners to rely on them. If a solo miner with 200 TH/s has a roughly one-in-four-million chance of finding any given block, and blocks are found approximately every ten minutes, that miner would wait an average of about 76 years between solo blocks. Pool mining converts this lottery into a steady paycheck. Over 95% of Bitcoin's hashrate is directed through mining pools.

Pool Centralization Concerns

As of early 2026, a small number of mining pools control a large share of the total hashrate. Foundry USA and AntPool together typically account for roughly half of global hashrate, and the four largest pools combined account for around three-quarters of it. This concentration raises legitimate security concerns: if a single entity controlled a majority of pooled hashrate, it could theoretically censor transactions or attempt double-spends.

The important distinction is that pool operators do not own the hashrate. Individual miners can switch pools at any time. If a pool operator behaved maliciously, its participants would migrate to competitors within hours. Pool centralization is a monitoring concern, not an existential threat. The Bitcoin community actively develops decentralized pool protocols (like Stratum V2) to reduce the trust miners must place in pool operators.

Energy and Bitcoin Mining

Bitcoin mining consumes a significant amount of energy. Estimates for the network's total power consumption range from 100 to 200+ TWh per year as of early 2026, comparable to the electricity consumption of a mid-sized country.

Energy Is the Security Budget

The energy consumed by miners is the direct cost of securing the Bitcoin network. Every kilowatt-hour spent mining becomes part of the cumulative work protecting the blockchain. To attack Bitcoin, an adversary must match or exceed this ongoing energy expenditure. The higher the energy consumption, the higher the cost of attack, and the more secure the network becomes.

The energy cost of securing money is not unique to Bitcoin. The global banking system consumes enormous energy across server farms, SWIFT infrastructure, armored transport, vaults, and regulatory apparatus, and Bitcoin achieves comparable security for a global monetary network with a fraction of that institutional overhead.

The "Waste" Argument

The claim that Bitcoin mining "wastes" energy assumes the energy would be better used elsewhere. This argument has several factual problems:

• Stranded energy: A significant portion of Bitcoin mining uses energy that has no other buyer. Flared natural gas at oil wells, curtailed wind and solar power, and hydroelectric surplus in remote locations are all energy sources that would otherwise be wasted entirely. Mining converts stranded energy into economic value.
• Grid balancing: Miners are among the most flexible large-scale energy consumers in the world. They can ramp down or shut off within seconds when grid demand spikes, selling their contracted energy back to the grid. In Texas, several large mining operations participate in demand response programs that help stabilize the grid during peak demand.
• Methane mitigation: Mining operations that capture and burn flared methane at oil wells reduce emissions of a greenhouse gas approximately 80 times more potent than CO2 over a 20-year horizon. These operations turn an environmental liability into a revenue source.

Renewable Energy and Mining

Miners are economically driven to find the cheapest available energy. Renewable sources, particularly hydroelectric, geothermal, wind, and solar, are increasingly the cheapest form of generation in many regions. Studies estimate that 50-60% of Bitcoin mining is powered by renewable or zero-carbon energy sources, a higher percentage than virtually any other major industry.

Mining also incentivizes new renewable energy development by serving as a "buyer of first resort" for new solar or wind installations, providing guaranteed revenue during the years before a facility connects to the broader grid.

Mining's Role in Bitcoin's Security

Bitcoin's security model rests on a simple principle: the cost of attacking the network must always exceed the potential profit from an attack. Mining is the mechanism that makes this true.

The 51% Attack

The most commonly discussed attack vector is the 51% attack: if a single entity controlled more than half the network's total hashrate, it could theoretically:

• Double-spend: Send bitcoin to someone, wait for confirmation, then rewrite the blockchain to redirect those same funds back to itself.
• Censor transactions: Refuse to include specific transactions in blocks.
• Prevent confirmations: Block other miners from having their blocks accepted by always building on its own chain.

What a 51% attacker cannot do is equally important:

• Create bitcoin out of thin air (violates consensus rules; nodes would reject the blocks).
• Steal bitcoin from other addresses (requires private keys; hashrate is irrelevant).
• Change the rules of the protocol (nodes enforce rules independently of miners).

The Economic Cost of Attack

Acquiring 51% of Bitcoin's hashrate would require:

• Hardware: At 800+ EH/s, an attacker needs over 400 EH/s of ASIC capacity. At current prices, this represents tens of billions of dollars in equipment alone, assuming it could be manufactured and delivered. ASIC production is capacity-constrained; fulfilling that demand quickly is physically impossible.
• Energy: Powering 400+ EH/s requires multiple gigawatts of continuous electrical power, demanding dedicated power plants or a significant fraction of a country's grid.
• Ongoing cost: The attacker must sustain this hashrate for the duration of the attack, spending millions of dollars per day on electricity alone.
• Self-defeating economics: A successful attack would crash the Bitcoin price, destroying the value of the attacker's own hardware and any bitcoin they hold. The attacker would need to profit more from the attack than they lose from the market crash.

The network does not rely on trust, reputation, or legal enforcement. It relies on the physical reality that energy is expensive and hashrate cannot be faked.

Why Geographic Distribution Matters

If mining were concentrated in a single country, that government could pressure miners to censor transactions, refuse to include certain addresses in blocks, or even attempt to reorganize the blockchain. Geographic distribution makes this kind of coordinated coercion far more difficult.

China's mining ban in June 2021 demonstrated both the risk of concentration and the network's resilience. Before the ban, an estimated 50-75% of Bitcoin's hashrate was located in China. When the government ordered all mining operations to shut down, the network's hashrate dropped by roughly half within weeks. Difficulty adjusted downward, blocks continued to be produced on schedule, and miners relocated to the US, Kazakhstan, Russia, and elsewhere. Within a year, total hashrate had recovered and surpassed its pre-ban level, now distributed across a much broader set of countries. Today, hashrate is more geographically distributed than at any point in Bitcoin's history, which strengthens the network against single-country regulatory risk.

Pool centralization is a separate concern from geographic distribution. A pool operator does not own the hashrate pointed at it, and miners can switch pools at any time, so a malicious operator would quickly lose participants. The more subtle risk is that pool operators select which transactions to include in candidate blocks and choose their order, which gives them room to prioritize certain transactions or pursue MEV-style (Miner Extractable Value) strategies. Protocols like Stratum V2 address this by letting individual miners construct their own block templates rather than accepting templates from the pool operator, shifting transaction-selection power back to the individual miner.

Security Beyond Hashrate

Mining provides the economic security layer, but Bitcoin's full security model includes additional protections: tens of thousands of full nodes independently verify every block (making it impossible for miners to change protocol rules), mining hardware is distributed across dozens of countries (making coordinated seizure effectively impossible), the protocol is open-source and publicly auditable, and the cryptographic foundations (ECDSA and Schnorr signatures, SHA-256, secp256k1) provide the mathematical security underlying every address, transaction, and block.

Frequently Asked Questions

Can you mine bitcoin on a regular computer?

Technically, yes. Any computer can run the SHA-256 algorithm. Practically, a regular computer's hashrate is so low compared to the network total that the expected time to find a block would be millions of years. Modern Bitcoin mining requires ASICs to be competitive. Even a top-end consumer GPU cannot produce meaningful hashrate compared to a single ASIC.

How long does it take to mine one bitcoin?

This depends on your share of the network hashrate. The network produces 3.125 bitcoin per block (set by the subsidy schedule in GetBlockSubsidy()), with a new block approximately every ten minutes (about 450 bitcoin per day across the entire network). A miner with 0.1% of the global hashrate would earn approximately 0.45 bitcoin per day on average. A single modern ASIC represents far less than 0.1% of the network and would take many years, well over a decade at current difficulty, to earn one full bitcoin.

What happens when all 21 million bitcoin have been mined?

The last bitcoin is projected to be mined around the year 2140. After that, miners will earn revenue solely from transaction fees. The transition is gradual: the block subsidy halves every four years, and transaction fees already constitute a meaningful portion of miner revenue during high-demand periods. The expectation is that as Bitcoin adoption grows, demand for block space (and therefore fees) will be sufficient to incentivize continued mining.

Is Bitcoin mining profitable?

Profitability depends on three variables: the Bitcoin price, electricity cost, and hardware efficiency. Miners with access to cheap electricity (below $0.05/kWh) and modern ASICs are generally profitable. Miners paying retail electricity rates with older hardware often operate at a loss. The difficulty adjustment ensures that profitability tends toward equilibrium: when mining is highly profitable, more miners join, increasing difficulty and reducing per-miner revenue until margins stabilize.

Does Bitcoin mining harm the environment?

Bitcoin mining consumes significant energy, but the environmental impact depends on the energy source. The Cambridge Centre for Alternative Finance estimated in its April 2025 report that 52.4% of Bitcoin mining draws on sustainable sources (42.6% renewables plus 9.8% nuclear), driven by miners' economic incentive to seek the cheapest power available. Mining operations also capture flared natural gas (reducing methane emissions), use stranded energy that would otherwise go unused, and participate in demand-response programs that help stabilize electrical grids.

Why does Bitcoin use proof-of-work instead of proof-of-stake?

Proof-of-work anchors Bitcoin's security in physical reality: energy expenditure and hardware investment that cannot be faked or duplicated. Proof-of-stake anchors security in the existing token distribution, creating a system where the wealthiest participants have the most control. Bitcoin's design prioritizes censorship resistance, decentralization, and security properties that are grounded in physics rather than game-theoretic assumptions about staker behavior. For a full comparison, see Proof of Work vs Proof of Stake.