How Bitcoin mining works
- Miners collect pending transactions from the mempool into a candidate block.
- They compute a SHA-256 hash of the block header + a nonce.
- If the hash is below the current difficulty target (starts with enough leading zeros), the block is valid.
- They broadcast the valid block; network accepts it; miner claims the block reward (currently 3.125 BTC as of April 2024 halving, halving again in ~2028) plus transaction fees.
- If not, they increment the nonce and try again — billions of times per second.
Difficulty adjusts every 2,016 blocks (~2 weeks) to keep average block time at 10 minutes regardless of total network hashrate.
Hardware evolution
- CPU mining (2009-2011) — anyone with a computer could mine. Modest hashrates.
- GPU mining (2011-2013) — consumer graphics cards outperformed CPUs ~10x. Home miners proliferated.
- FPGA mining (2012-2013) — brief intermediate step.
- ASIC mining (2013-present) — application-specific integrated circuits designed for SHA-256 only. Current generation (Bitmain S21, Whatsminer M60) deliver 200+ TH/s at ~15 J/TH efficiency. Home mining is economically dead for Bitcoin.
Other PoW coins still have GPU-accessible algorithms — Kaspa (GhostDAG), Ergo, some privacy coins — where home mining remains relevant.
Mining economics
Miner revenue = block subsidy + transaction fees. Costs = electricity + hardware depreciation + facility overhead.
Profitability depends on:
- Electricity cost — the dominant variable input. Industrial rates of $0.04-0.06/kWh separate profitable miners from unprofitable ones.
- Hashrate share — your share of daily global rewards proportional to your share of total network hashrate.
- ASIC efficiency — newer hardware does more TH per watt. Miners running obsolete hardware get priced out first during tough markets.
- Network difficulty — rises with total hashrate, reducing per-unit rewards.
- Bitcoin price — the multiplier on all other inputs.
Publicly-traded miners (Marathon, Riot, CleanSpark, Core Scientific, Hut 8) offer retail equity exposure to mining economics. Stock returns are highly correlated with BTC price but amplified by operating leverage.
Geographic dynamics
After China banned mining in May 2021, industrial mining redistributed:
- United States — now ~35-40% of global hashrate. Texas, Kentucky, Georgia lead.
- Russia and Kazakhstan — ~12-15% combined.
- Rest of world — the remainder spread across Canada, Paraguay, UAE, others.
Mining clusters near cheap stranded energy: hydro (Norway, Iceland, Washington state), natural gas flaring sites (Texas, North Dakota), and coal/nuclear where electricity prices are low.
Risks and considerations
- Halving cycles — every ~4 years, block subsidy halves. Marginal miners become unprofitable unless BTC price rises or efficiency improves. Post-halving survival is the miner’s core challenge.
- Regulatory risk — governments can (and do) ban or restrict mining. Texas grid stress, EU energy scrutiny, and China’s blanket ban all demonstrate.
- Equipment supply chain — ASIC production is concentrated (Bitmain, MicroBT). Shortages and tariffs are a structural risk.
- Long-term security budget — as block subsidy approaches zero, fees must sustain miner revenue. Whether fees scale sufficiently is a long-running debate.
For investors, mining stocks offer leveraged BTC exposure with operational risk. For operators, the game is increasingly concentrated among large industrial players with access to cheap power and efficient hardware.