Definition
Network hashrate is the total computational power that all Bitcoin miners are collectively pointing at the network at any given moment, expressed as the number of SHA-256 double-hashes performed per second. It is the aggregate of every working ASIC on Earth, from a single Bitaxe in a closet to multi-megawatt Hashcenters, and it is the single best proxy for how secure and how competitive the Bitcoin network is.
Network hashrate cannot be measured directly — no one polls every miner. Instead it is estimated by observing how fast blocks are actually being found relative to the current difficulty. If blocks arrive faster than the targeted 10-minute average, the network is assumed to be running more hashrate than the difficulty was calibrated for; if they arrive slower, hashrate has dropped. As of 2024–2025, Bitcoin’s estimated network hashrate regularly exceeds 600 EH/s (600 exahashes, or 600 × 1018 hashes per second).
How it actually works under the hood
Every hash an ASIC computes is a lottery ticket. A miner takes the block header, hashes it twice with SHA-256, and checks whether the result is numerically less than or equal to a proof-of-work target. The protocol expresses that target through difficulty using the relationship target = 2^256 / (65536 × difficulty) — a higher difficulty means a smaller target, which means each individual hash is less likely to win. Network hashrate is, in effect, how many of these lottery tickets the entire planet is buying per second.
Because the protocol wants a block roughly every 10 minutes regardless of how much hardware is online, it re-tunes the target through difficulty adjustment every 2,016 blocks (about two weeks). Rising network hashrate pushes blocks out faster, which forces difficulty up at the next retarget; falling hashrate does the opposite. Hashrate and difficulty therefore chase each other in a feedback loop, with difficulty acting as the slow-moving average of the much noisier real-time hashrate estimate.
Why it matters for your machines
For anyone running real hardware, network hashrate is the denominator in your share of the rewards. Your odds of contributing to a found block are roughly your own hashrate divided by the network hashrate. When global hashrate climbs and your fleet stays the same size, your slice of every block reward shrinks — the same watts now earn fewer sats. This is why miners obsess over efficiency (joules per terahash) rather than raw hashrate alone: in a rising-hashrate world, only the most efficient machines stay profitable.
This dynamic also shapes hardware and firmware decisions. As network hashrate grows, older units like the S9 or early S17 generations get squeezed out first, while newer-generation ASICs with better efficiency keep their margin longer. It is also why tuning and firmware choices matter: getting more usable hashrate per watt out of an existing machine effectively buys you more time before the rising network curve makes a unit uneconomical. The practical levers most miners reach for are:
- Efficiency tuning — underclocking or undervolting to lower joules per terahash, trading a little hashrate for a lot of margin.
- Pool participation — joining a mining pool so a small share of network hashrate still produces steady payouts instead of all-or-nothing solo luck.
- Uptime and cooling — every minute offline is hashrate you contributed to nobody, and thermal throttling quietly erodes your real contribution.
It is worth separating network hashrate from your own reported hashrate. Your miner’s dashboard shows the local figure; the network figure is the global estimate you are competing against. The gap between the two — and how it trends over the months you own a machine — is what ultimately decides whether that hardware keeps paying for its electricity.
Network hashrate distribution across pools is itself a decentralization concern. When a handful of large pools command most of the global hashrate, block construction concentrates in few hands — one of the reasons protocols like Stratum V2 and template-construction approaches matter, since they push block-building back toward individual operators. Running your own efficient hardware, even a single unit, adds one more sliver of independent hashrate to the network.
If you are sizing a fleet, comparing efficiency across generations, or deciding which machine still earns at today’s difficulty, start with the current-generation ASICs in our miner catalog — and if a unit is underperforming its rated hashrate, the ASIC troubleshooting guides will help you recover the hashrate you are paying watts for.
In Simple Terms
The total computing power of all Bitcoin miners combined. Higher network hashrate means more competition.
Network hashrate is the total computational power that all Bitcoin miners are collectively pointing at the network at any given moment, expressed as the number of SHA-256 double-hashes performed per second. It is the aggregate of every working ASIC on Earth, from a single Bitaxe in a closet to multi-megawatt Hashcenters, and it is the single best proxy for how secure and how competitive the Bitcoin network is.
Network hashrate cannot be measured directly — no one polls every miner. Instead it is estimated by observing how fast blocks are actually being found relative to the current difficulty. If blocks arrive faster than the targeted 10-minute average, the network is assumed to be running more hashrate than the difficulty was calibrated for; if they arrive slower, hashrate has dropped. As of 2024–2025, Bitcoin’s estimated network hashrate regularly exceeds 600 EH/s (600 exahashes, or 600 × 1018 hashes per second).
How it actually works under the hood
Every hash an ASIC computes is a lottery ticket. A miner takes the block header, hashes it twice with SHA-256, and checks whether the result is numerically less than or equal to a proof-of-work target. The protocol expresses that target through difficulty using the relationship target = 2^256 / (65536 × difficulty) — a higher difficulty means a smaller target, which means each individual hash is less likely to win. Network hashrate is, in effect, how many of these lottery tickets the entire planet is buying per second.
Because the protocol wants a block roughly every 10 minutes regardless of how much hardware is online, it re-tunes the target through difficulty adjustment every 2,016 blocks (about two weeks). Rising network hashrate pushes blocks out faster, which forces difficulty up at the next retarget; falling hashrate does the opposite. Hashrate and difficulty therefore chase each other in a feedback loop, with difficulty acting as the slow-moving average of the much noisier real-time hashrate estimate.
Why it matters for your machines
For anyone running real hardware, network hashrate is the denominator in your share of the rewards. Your odds of contributing to a found block are roughly your own hashrate divided by the network hashrate. When global hashrate climbs and your fleet stays the same size, your slice of every block reward shrinks — the same watts now earn fewer sats. This is why miners obsess over efficiency (joules per terahash) rather than raw hashrate alone: in a rising-hashrate world, only the most efficient machines stay profitable.
This dynamic also shapes hardware and firmware decisions. As network hashrate grows, older units like the S9 or early S17 generations get squeezed out first, while newer-generation ASICs with better efficiency keep their margin longer. It is also why tuning and firmware choices matter: getting more usable hashrate per watt out of an existing machine effectively buys you more time before the rising network curve makes a unit uneconomical. The practical levers most miners reach for are:
- Efficiency tuning — underclocking or undervolting to lower joules per terahash, trading a little hashrate for a lot of margin.
- Pool participation — joining a mining pool so a small share of network hashrate still produces steady payouts instead of all-or-nothing solo luck.
- Uptime and cooling — every minute offline is hashrate you contributed to nobody, and thermal throttling quietly erodes your real contribution.
It is worth separating network hashrate from your own reported hashrate. Your miner’s dashboard shows the local figure; the network figure is the global estimate you are competing against. The gap between the two — and how it trends over the months you own a machine — is what ultimately decides whether that hardware keeps paying for its electricity.
Network hashrate distribution across pools is itself a decentralization concern. When a handful of large pools command most of the global hashrate, block construction concentrates in few hands — one of the reasons protocols like Stratum V2 and template-construction approaches matter, since they push block-building back toward individual operators. Running your own efficient hardware, even a single unit, adds one more sliver of independent hashrate to the network.
If you are sizing a fleet, comparing efficiency across generations, or deciding which machine still earns at today’s difficulty, start with the current-generation ASICs in our miner catalog — and if a unit is underperforming its rated hashrate, the ASIC troubleshooting guides will help you recover the hashrate you are paying watts for.