What is hash rate?

The combined speed at which miners try candidate block hashes, measured in hashes per second and usually quoted in exahashes per second.

Hash rate is not published by any single authority; it is estimated from how fast blocks are found against the current difficulty, which is why different sites report slightly different figures for the same day. A rising hash rate means an attacker needs more equipment to rewrite recent history, so it is a rough measure of how expensive a chain reorganization would be.

It says nothing about price, and no exchange availability decision depends on it. Individual miners rarely solve blocks alone at these speeds, which is why most join a mining pool. The mistake to avoid is reading a hash rate chart as a demand signal. It measures how much equipment is switched on, which follows revenue and electricity prices with a lag.

How it works

No one counts hashes. Every attempt that fails leaves no trace anywhere, so the network rate is inferred from the only two things that are observable: the difficulty currently in force and how quickly blocks are actually arriving.

Bitcoin targets one block every ten minutes and adjusts difficulty every 2,016 blocks, which is about two weeks, so that the observed rate returns to that target (source: the Bitcoin protocol's consensus rules). If blocks came faster than ten minutes on average during the period, difficulty rises; slower, and it falls. An estimator inverts that relationship: given the difficulty and the observed interval between blocks, it solves for the rate that would produce those intervals.

Two consequences follow. First, the estimate is noisy over short windows, because block discovery is a random process and a quiet hour is ordinary rather than meaningful. Sites smooth over different windows, which is why their numbers differ. Second, the figure lags reality by up to a full adjustment period after a large change in deployed equipment.

Units climb in thousands: a terahash is a trillion hashes per second, a petahash a thousand of those, and an exahash a thousand petahashes.

Example

Illustrative arithmetic. Suppose a network's blocks arrived every 9 minutes on average over an adjustment period rather than the 10 minutes targeted. Miners produced blocks about 11 percent faster than intended, so the next adjustment raises difficulty by roughly the same 11 percent, and the average interval returns to 10 minutes at the new level. Nothing about the coin changed. What changed is how much equipment was pointed at it, and the protocol absorbed that automatically.

Why it matters when you buy

Hash rate belongs to a network's security story, not to a buying decision, and it does not affect what you pay or where you can trade. What does affect both is the market structure around the asset: the liquidity view shows measured spreads and depth per venue, and the concentration view shows how much of an asset's volume sits on a single exchange.

Related terms

• hash: the function miners are computing
• difficulty adjustment: the mechanism that keeps block timing steady
• mining pool: how small miners combine their output
• proof of work: the consensus method hash rate secures
• chain reorganization: what a large share of hash rate could attempt
• block reward: the revenue that draws equipment in

Questions

Why do two sites report different hash rates for the same day?

Both are estimating from block intervals, and they smooth over different windows. Neither is measuring hashes directly, so a few percent of disagreement is normal.

Does a falling hash rate mean the network is in trouble?

Not by itself. Equipment goes offline for seasonal power prices, weather, and hardware upgrades, and the next difficulty adjustment restores the target block interval regardless.

Does hash rate affect proof-of-stake chains?

No. Chains secured by staked capital have no mining, so the equivalent question is how much value is staked and how it is distributed across validators.