When Satoshi Nakamoto mined the first Bitcoin block on January 3, 2009, the entire operation ran on a single CPU pulling roughly 4.7 MH/s. Fast forward to 2026, and the Bitcoin network’s combined hash rate regularly exceeds 800 EH/s, a figure that represents a trillion-fold increase in computational power dedicated to securing the world’s largest cryptocurrency. That trajectory, from a lone laptop to continent-spanning industrial facilities, tells one of the most remarkable infrastructure stories in modern finance. Along the way, mining has reshaped energy markets, triggered geopolitical shifts, and turned specialized silicon chips into some of the most sought-after hardware on the planet.
Key Takeaways
- Bitcoin mining began in 2009 using ordinary CPUs, with miners earning 50 BTC per block at virtually zero competition.
- GPU mining (2010) delivered roughly 100x the hash rate of CPUs, sparking the first mining arms race.
- ASIC miners, introduced in 2013, made all previous hardware obsolete and industrialized the mining sector.
- China’s mining ban in June 2021 displaced over 50% of global hash rate, triggering the largest geographic redistribution in Bitcoin’s history.
- The April 2024 halving cut the block reward to 3.125 BTC, pushing miners toward transaction fee revenue and energy efficiency.
- As of early 2026, the United States leads global hash rate share at approximately 38%, followed by Russia and Kazakhstan.
The CPU Era: Mining on a Laptop (2009 to 2010)
Bitcoin’s earliest days were remarkably low-tech. Nakamoto’s original client included a built-in mining function, and anyone with a standard desktop computer could participate. Block difficulty hovered near 1.0 for the first several months, meaning a single Intel or AMD processor could solve a block in a matter of hours. The reward was generous: 50 BTC per block, worth essentially nothing at the time but later valued at millions of dollars.
The first known commercial Bitcoin transaction, Laszlo Hanyecz’s famous purchase of two pizzas for 10,000 BTC on May 22, 2010, happened squarely within the CPU mining era. Hanyecz himself was one of the first people to experiment with GPU mining, a move that would soon make CPUs obsolete for Bitcoin mining entirely. For a deeper look at Bitcoin’s founding period and early milestones, see our history of Bitcoin overview.
During this period, mining was essentially free. Electricity costs were negligible relative to the computational effort required, and there were no specialized facilities. Miners ran the software alongside their everyday applications, treating it more as an experiment than a business.
The GPU and FPGA Revolution (2010 to 2013)
By late 2010, miners discovered that graphics processing units (GPUs) could execute Bitcoin’s SHA-256 hashing algorithm far more efficiently than CPUs. A single high-end GPU could deliver 100 MH/s or more, compared to roughly 1 to 10 MH/s from a CPU. This shift turned mining into a competitive pursuit for the first time, with hobbyists building multi-GPU rigs in basements and garages.
AMD’s Radeon HD 5870 and its successors became the de facto mining cards, prized for their parallel processing architecture. Mining pools, which allowed multiple participants to combine hash power and split rewards, also emerged during this period. Slush Pool (now Braiins Pool) launched in November 2010 as the world’s first mining pool, fundamentally changing the economics of small-scale mining.
A brief but notable interlude came with field-programmable gate arrays (FPGAs), which offered better energy efficiency than GPUs without the cost of custom chip fabrication. Companies like Butterfly Labs sold FPGA mining boards starting in 2011. However, FPGAs occupied the market for barely two years before ASICs arrived and rendered them obsolete.
The ASIC Era and Industrialization (2013 to 2020)
Application-specific integrated circuits (ASICs) represent hardware designed to do one thing: mine Bitcoin. When Canaan Creative shipped the Avalon ASIC miner in January 2013, it marked a turning point. Within months, Bitmain’s Antminer series followed, and the mining landscape changed permanently. Home miners running GPUs could no longer compete with purpose-built machines delivering terahashes per second at a fraction of the energy cost.
The industrialization that followed was swift. Mining migrated from spare bedrooms to dedicated warehouses, then to massive data centers in regions with cheap electricity. Inner Mongolia, Sichuan, and Xinjiang in China became global mining hubs, leveraging coal power and seasonal hydroelectric surplus. By 2017, an estimated 65% to 75% of global hash rate resided within China’s borders.
Each new ASIC generation roughly doubled energy efficiency while increasing hash output. The competitive pressure forced continuous capital reinvestment, squeezing out smaller operators and consolidating the industry around a handful of large manufacturers and mining firms.
| Era | Typical Hardware | Hash Rate | Approx. Hash/Dollar | Years Active |
|---|---|---|---|---|
| CPU | Intel Core i7 | 1 to 10 MH/s | ~2 MH/$ | 2009 to 2010 |
| GPU | AMD Radeon HD 5870 | 100 to 400 MH/s | ~50 MH/$ | 2010 to 2013 |
| FPGA | Butterfly Labs Single | 800 MH/s | ~130 MH/$ | 2011 to 2013 |
| ASIC (Early) | Bitmain Antminer S1 | 180 GH/s | ~600 MH/$ | 2013 to 2016 |
| ASIC (Modern) | Antminer S21 Hydro | 335 TH/s | ~50 GH/$ | 2023 to present |
Source: Manufacturer specifications, CoinDesk mining hardware archives
The table above illustrates the staggering improvements in mining efficiency across hardware generations. Each leap made previous equipment functionally worthless within months of a new generation’s release, a pattern that continues today.
China’s Dominance and the Great Migration (2017 to 2022)
For the better part of a decade, China was the undisputed center of Bitcoin mining. Cheap electricity (particularly hydropower in Sichuan during the wet season), proximity to ASIC manufacturers, and relatively permissive local policies created ideal conditions. At the peak in early 2021, the Cambridge Centre for Alternative Finance (CCAF) estimated that China controlled approximately 46% of global hash rate, down from a high near 75% in 2019.
That dominance ended abruptly. In May 2021, China’s State Council called for a crackdown on Bitcoin mining and trading. Provincial governments in Inner Mongolia, Yunnan, Sichuan, and Xinjiang issued shutdown orders throughout June and July 2021. The effect was immediate and dramatic: Bitcoin’s network hash rate dropped roughly 50% between May and July 2021, falling from approximately 180 EH/s to under 90 EH/s.
What followed was the largest geographic redistribution of computing power in history. We’ve tracked this migration closely in our coverage, and the speed at which miners relocated equipment across continents remains one of the most underappreciated logistics stories in the crypto industry. Hundreds of thousands of ASIC units were shipped to the United States, Kazakhstan, Russia, and Canada within months.
| Country | Hash Rate Share (Jan 2021) | Hash Rate Share (Jan 2022) | Hash Rate Share (Jan 2026 Est.) |
|---|---|---|---|
| China | 46% | 21% | ~15% |
| United States | 17% | 38% | ~38% |
| Kazakhstan | 8% | 13% | ~6% |
| Russia | 7% | 11% | ~12% |
| Canada | 5% | 7% | ~7% |
| Other | 17% | 10% | ~22% |
Source: Cambridge Centre for Alternative Finance (CCAF), 2026 estimates based on CCAF methodology and public miner disclosures
The Great Migration had lasting consequences. The United States rapidly became the world’s dominant mining nation, with publicly traded companies like Marathon Digital, Riot Platforms, and CleanSpark scaling operations across Texas, Georgia, and North Dakota. Kazakhstan initially absorbed significant hash rate but subsequently imposed its own restrictions, including an electricity surcharge on miners, driving some operators to relocate again.
China’s share, notably, did not fall to zero. Underground mining operations persisted, and CCAF data suggests a residual hash rate share of roughly 15% to 20%, operating through VPNs and proxy pools. This persistence highlights the difficulty of enforcing outright bans on a decentralized network.
The Energy Debate: Waste or Innovation?
No discussion of Bitcoin mining history is complete without addressing the energy controversy. Bitcoin’s annualized electricity consumption, estimated by the CCAF at approximately 150 TWh in 2025, places it in the range of some mid-sized countries. Critics argue this represents an unacceptable environmental cost for a financial network. Proponents counter that mining incentivizes renewable energy development and utilizes stranded energy that would otherwise go to waste.
The data tells a nuanced story. According to the Bitcoin Mining Council’s surveys, the sustainable energy mix for Bitcoin mining reached approximately 59.5% by Q4 2024, up from 36% in 2021. Miners have increasingly sought out flared natural gas at oil wells, curtailed wind and solar output, and geothermal sources in countries like El Salvador and Iceland.
Texas provides a compelling case study. Miners there participate in demand-response programs with ERCOT (the state’s grid operator), powering down during peak demand periods and earning credits that improve their economics. This arrangement effectively turns miners into flexible load resources, a concept that energy planners have started to take seriously. Our coverage shows that this model is now being replicated in several other U.S. states and parts of northern Europe.
| Year | Estimated Network Power (TWh) | Sustainable Energy Mix | Network Hash Rate (EH/s) |
|---|---|---|---|
| 2017 | ~15 | ~28% | ~12 |
| 2019 | ~60 | ~30% | ~90 |
| 2021 | ~90 | ~36% | ~180 |
| 2023 | ~120 | ~53% | ~450 |
| 2025 | ~150 | ~60% | ~750 |
| 2026 (est.) | ~155 | ~62% | ~800+ |
Source: Cambridge Centre for Alternative Finance, Bitcoin Mining Council quarterly surveys, Hashrate Index
The Halving Cycles and Mining Economics
Bitcoin’s programmatic halving events, which cut the block reward in half roughly every four years, have shaped mining economics at every stage. The halvings create a predictable supply shock that historically correlates with price appreciation, though the margin compression they impose on miners is equally significant.
The first halving in November 2012 reduced the reward from 50 to 25 BTC. At the time, Bitcoin traded near $12, and the event barely registered outside niche forums. The second halving in July 2016 (25 to 12.5 BTC) coincided with Bitcoin’s rise from $650 to nearly $20,000 by late 2017. The third halving in May 2020 (12.5 to 6.25 BTC) preceded the bull run that took Bitcoin above $69,000 in November 2021.
The most recent halving occurred in April 2024, cutting the reward to 3.125 BTC per block. This event forced another wave of efficiency upgrades across the industry. Miners operating older-generation ASICs with energy costs above approximately $0.07 per kWh found themselves unprofitable almost overnight. The result was a consolidation wave, with larger, well-capitalized firms acquiring distressed operations at steep discounts.
Transaction fees have become an increasingly important revenue component. The introduction of Ordinals and BRC-20 tokens in 2023 briefly pushed transaction fees above block rewards during peak periods. While that intensity has not sustained, it demonstrated that a fee-based security model can function, a critical question as Bitcoin approaches its 21 million coin supply cap. For context on how these dynamics affect the broader exchange ecosystem, see our crypto exchange statistics analysis.
Bitcoin Mining in 2026: The Current State
As of early 2026, Bitcoin mining has matured into a sophisticated, publicly traded industry sector. The network hash rate sits above 800 EH/s, a figure that was almost unimaginable five years ago. The latest generation of ASIC miners, including Bitmain’s Antminer S21 series and MicroBT’s WhatsMiner M60 line, deliver over 300 TH/s per unit while consuming under 20 joules per terahash.
The industry’s geographic distribution has stabilized somewhat since the post-China scramble. The United States remains dominant, with Texas, Georgia, and New York hosting the largest concentrations of hash power. Russia has grown its share to roughly 12%, aided by cheap natural gas and a regulatory environment that, while ambiguous, has not actively suppressed mining. Emerging markets in the Middle East (particularly the UAE and Oman) and Latin America (Paraguay and Argentina) are attracting new facilities with low energy costs and favorable climate conditions.
Mining pool concentration remains a concern for network decentralization. Foundry USA and AntPool together regularly account for over 50% of blocks mined, though it is worth noting that pool participants can switch allegiance quickly, and the trend toward Stratum V2 protocol adoption gives individual miners more control over block template construction.
Perhaps the most significant development in 2025 and 2026 has been the convergence of Bitcoin mining with artificial intelligence infrastructure. Several major mining companies, including Core Scientific and Hut 8, have begun converting portions of their facilities to host AI computing workloads, leveraging their existing power infrastructure and cooling systems. This diversification strategy hedges against Bitcoin price volatility while capitalizing on the explosive demand for GPU compute capacity.
Frequently Asked Questions (FAQs)
Bitcoin mining started on January 3, 2009, when Satoshi Nakamoto mined the genesis block (Block 0) using a standard CPU. For the first year, mining required only a basic desktop computer, and the block reward was 50 BTC. The network difficulty was so low that blocks could be solved in minutes with consumer hardware.
No, mining Bitcoin with a regular computer is no longer viable. The network difficulty has increased to a level where only ASIC miners, purpose-built machines costing thousands of dollars, can compete. A modern CPU would take millions of years to solve a single block. Home miners can participate through mining pools using ASICs, but standalone CPU or GPU mining yields negligible returns.
China banned Bitcoin mining in 2021 primarily due to concerns about financial stability, energy consumption, and carbon emission targets. The State Council cited risks associated with speculative trading and the strain mining placed on provincial power grids, particularly coal-dependent ones. The ban displaced roughly half of the global hash rate within weeks.
Bitcoin mining consumes approximately 150 to 155 TWh of electricity annually as of 2025 to 2026, roughly comparable to a country like Poland. About 60% of this energy comes from sustainable sources, including hydropower, wind, solar, and nuclear. The energy mix has improved significantly since 2021, when the sustainable share was closer to 36%.
When all 21 million Bitcoins are mined (estimated around 2140), miners will rely entirely on transaction fees for revenue. This transition is already underway, as halving events progressively reduce block rewards. The network’s security will depend on sufficient fee volume to incentivize miners, a model that Ordinals and BRC-20 token activity in 2023 showed can generate substantial fee revenue.
Conclusion
The history of Bitcoin mining is a story of relentless optimization. From a single CPU in 2009 to an 800+ EH/s global network in 2026, every phase has been defined by the pursuit of greater efficiency, whether through hardware innovation, geographic arbitrage, or energy sourcing. China’s ban proved that no single nation controls Bitcoin’s infrastructure, and the Great Migration demonstrated the network’s resilience in the face of regulatory shock.
Looking ahead, the economics will only tighten. Future halvings will continue compressing block rewards, forcing miners to operate at ever-narrower margins or diversify into adjacent computing markets. The energy debate will persist, though the trend toward sustainable sources and grid-stabilization partnerships offers a credible path forward. What remains constant is the core incentive structure Nakamoto designed: as long as securing the network pays, miners will find a way to do it profitably.
