Bitcoin Mining · Chart

Bitcoin hashrate chart (EH/s) with 30 / 60 / 200-day moving averages

Total Bitcoin network hashrate, in exahash per second, plotted on a log scale with 30, 60 and 200-day moving averages. Hashrate is the most direct measure of network security and miner-economic confidence: rising hashrate means new mining capacity coming online, falling hashrate means miners switching off or relocating. The 30-day MA captures short-term miner sentiment; the 200-day MA defines the long-run hashrate trend; the 30-vs-200 spread is a tell on whether the network is expanding or in capitulation. Pair with the difficulty chart and Hash Ribbons to triangulate where the mining economy sits in the cycle.

Last updated: 27 September 2026 By Govind Satoshi

Chart

Log-scale Bitcoin network hashrate in EH/s from 2014 onwards, with 30-day (green), 60-day (yellow) and 200-day (red) moving averages. The 200-day MA is the long-run trend; deviations of the 30-day MA above or below it mark expansion or capitulation. Hover any point for the exact hashrate and MA values on that day.

Loading data...

What is Bitcoin hashrate?

Hashrate is the total number of SHA-256 hash computations the Bitcoin mining network performs per second. Every mining ASIC tries trillions of nonce values per second, hashing each candidate block header until one of the hashes lands below the current difficulty target. Sum the per-ASIC hashrate across every active miner globally and you have the network hashrate.

The number is enormous. As of 8 October 2026, the 30-day average is about 958 EH/s, after a daily peak estimate of 1,306 EH/s on 25 October 2025. To put that in scale:

  • 1 TH/s = 10^12 hashes per second. A modern ASIC like the Antminer S21 does ~200 TH/s.
  • 1 PH/s = 10^15 hashes per second. A small mining operation with 5,000 S21s.
  • 1 EH/s = 10^18 hashes per second. A major industrial mining campus.
  • ~958 EH/s = current network total (30-day average). Roughly 4.8 million S21-equivalent ASICs running 24/7.

Hashrate is the direct measure of two things at once: network security (51 percent attack difficulty scales with hashrate) and miner-capital deployed (more hashrate = more electricity bought + more ASIC capex amortising).

Reading the moving averages

The chart plots the daily hashrate estimate (faint grey) with three moving averages overlaid:

  • 30-day MA (green). Short-term miner sentiment. Captures fleet on/off decisions over the past month. Responds to electricity price shocks, regulatory events (China 2021, Kazakhstan 2022 grid disruptions), and quarterly ASIC deployment waves.
  • 60-day MA (yellow). Medium-term trend. Smooths out single-event volatility while still responding within a quarter to genuine fleet-size changes.
  • 200-day MA (red). Long-run trend. The reference line for "is the network expanding or contracting structurally". Crossings of the 30-day above the 200-day mark expansion; below mark capitulation.

The Hash Ribbons indicator (separate tool) formalises this: when the 30-day crosses the 60-day from below after a period of capitulation, it has historically signalled a miner-capitulation bottom and a price-recovery setup. See the Hash Ribbons chart for the signal-state classifier and historical buy markers.

Hashrate cycle context

Hashrate growth has tracked Bitcoin's price cycles loosely, with these notable inflection points:

  • 2014-2015. Hashrate climbed from about 10 PH/s at the start of 2014 to about 300 PH/s by the end of 2014 and about 700 PH/s by the end of 2015, as the first generations of ASICs (Bitmain Antminer S1 to S5) displaced GPU and FPGA mining entirely.
  • 2017-2018. The 30-day average rose about sixfold during the 2017 bull market, from roughly 2 EH/s to 13 EH/s, and kept climbing to about 52 EH/s in October 2018. The 2018 bear then forced a major capitulation: the 30-day average fell about 28 percent to roughly 38 EH/s by late December 2018 as high-cost miners switched off.
  • May-July 2021. The China mining ban (the largest drawdown in Bitcoin's history): the 7-day average fell from about 180 EH/s in mid-May to about 90 EH/s in early July, roughly half the network in about seven weeks, and difficulty fell 28 percent in one adjustment. Mining capacity relocated to the US (Texas), Kazakhstan, and Russia, and difficulty was back at its May 2021 level by January 2022.
  • 2022 bear market. Hashrate held up much better than in prior bears because the 2021 capitulation had already shaken out high-cost miners. Modest pullbacks during the FTX collapse and the November 2022 difficulty trough.
  • 2024 halving (April 2024). The block subsidy halved from 6.25 to 3.125 BTC. Higher-cost ASICs (S19 and older) became unprofitable; some capitulated, replaced by newer S21-class fleets. Net hashrate growth resumed within months as electricity-cost-leader miners absorbed market share.
  • 2025-2026. Hashrate climbed to a daily peak estimate of 1,306 EH/s on 25 October 2025 on industrial-scale fleet expansion (Cipher, Riot, Marathon, CleanSpark, Iris Energy) and renewable-energy deployments. As BTC fell in 2026, difficulty was cut by 11 percent in February and 10 percent in June as miners switched off. The 30-day average is 958 EH/s against a 200-day average of 942 EH/s (above trend).

Methodology

  1. Inputs. Daily Bitcoin network hashrate (EH/s) from /assets/data/btc-hashrate.json. The full history from January 2014 comes from mempool.space's daily network hashrate estimate (blockchain.info as fallback), re-fetched on every automatic daily refresh. Each refresh is checked against known reference points, including the mid-2021 China ban drawdown, so no part of the series is interpolated. Daily values are estimates derived from block times and are noisy, which is why the moving averages matter.
  2. Moving averages. 30-day, 60-day and 200-day simple moving averages computed in-browser from the daily series. MAs are unavailable for the first N-1 days of the series (warm-up period).
  3. YoY change. Current hashrate vs hashrate exactly 365 days earlier, expressed as a percentage.
  4. Trend classification. Expansion if 30D MA > 200D MA × 1.05; above trend if 30D MA > 200D MA; near trend if 30D MA > 200D MA × 0.95; capitulation otherwise.
  5. Static-first. If the upstream API is unreachable, the existing JSON file is preserved and the build continues. The chart always renders against the last-known-good dataset.

Frequently asked questions

Hashrate is the total number of cryptographic hash computations performed by the Bitcoin network per second. It is the direct measure of computational power being thrown at the SHA-256 mining puzzle. Higher hashrate means more security (the network is harder to 51-percent attack) and more miner-capital deployed (rising hashrate signals miner confidence in forward profitability). Hashrate is expressed in hashes per second; the 30-day average was about 958 EH/s as of 8 October 2026 (exahash, where 1 EH/s = 10^18 hashes per second).

Raw daily hashrate is a noisy estimator. It is calculated from the time between blocks, which is itself stochastic (target 10 minutes but actual block-times range from seconds to over an hour). Daily values can swing 5-10 percent purely from block-time variance, not real hashrate change. The 30-day MA smooths out that block-time noise. The 60-day MA captures medium-term miner trends. The 200-day MA is the long-run reference line. The chart plots all three so you can see short-term shifts against the broader trend without being fooled by single-day spikes.

Hashrate and price are correlated long-term but the causality is debated. The mainstream view: price drives hashrate. Higher price = higher revenue per block = more profitable mining = more miners come online = hashrate rises. There is no mechanical reason for hashrate to push price up; new ASICs do not buy BTC, they just take a bigger share of block rewards. The contrarian view (Capriole Hash Ribbons thesis): when hashrate recovers after a deep miner capitulation (Hash Ribbons buy signal), the price often follows because the worst sellers (capitulating miners dumping inventory) have exhausted. Treat hashrate as a confirmation indicator, not a leading price oracle.

Hashrate is the actual computational power on the network at any given moment, estimated from observed block-times. Difficulty is a fixed parameter the protocol adjusts every 2016 blocks (~2 weeks) to target a 10-minute average block-time. Difficulty is a lagging indicator of hashrate: if hashrate rises, blocks come faster than 10 minutes, and the next adjustment raises difficulty proportionally. The difficulty chart is the auditable, on-chain record; the hashrate chart is the higher-resolution but slightly noisier real-time estimate. Both metrics tell the same story over months, but hashrate moves first.

The May-July 2021 China mining ban was the largest hashrate drawdown in Bitcoin's history. China at the time hosted roughly 50-65 percent of global mining (concentrated in Sichuan and Xinjiang) and Beijing's regulatory crackdown forced the largest fleets offline almost overnight. Published estimates put the fall at roughly half of network hashrate between mid-May and early July 2021. Difficulty, the on-chain record, fell 28 percent in a single adjustment on 3 July 2021 and did not regain its May 2021 level until January 2022, as fleets relocated to the US, Kazakhstan and elsewhere. On this chart's daily data the 7-day average fell from about 180 EH/s around 12 May 2021 to about 90 EH/s in early July. The Hash Ribbons chart marks a buy signal on 9 August 2021, when the 30-day average climbed back above the 60-day.

The estimate is statistical, not direct. Bitcoin protocol does not broadcast 'how much hashrate exists' as a field anywhere. Hashrate is inferred from the rate at which valid blocks are produced compared to the network's current difficulty target. The inference is unbiased over reasonable windows but noisy over short windows. The 30-day MA on this chart is the lowest-noise short-term estimate that is still responsive; longer windows lose responsiveness but gain smoothness. Mempool.space, blockchain.info, Coin Metrics and Glassnode all produce slightly different daily values from the same underlying block data because they use different smoothing windows.

About the author

Govind Satoshi
Govind Satoshi
Former Institutional Trader. Founder, SatoshiMacro.
Traded allocated institutional capital at a Sydney proprietary trading firm.