How network difficulty affects mining income

Difficulty retargets every 2016 blocks, roughly every two weeks. The 19.09.2026 adjustment came in at +4.16%, pushing difficulty from 127.45 T to 132.76 T (Hashrate Index, 21.09.2026). Hashprice on difficulty alone dropped from $39.36 (31.08.2026) to about $37.8 per PH per day, though the BTC price brought spot hashprice back to $40.93. The relationship between difficulty and income is inverse and not one for one: a 5% difficulty increase takes 4.76% of revenue. Net profit after electricity drops far harder, because your power bill does not retarget along with the network. This page covers the mechanics, the math, and what to do with both.

POOL BTC is an independent comparison site. We do not accept hashrate, pay rewards, or sell hardware.

What is Bitcoin network difficulty?

Difficulty is the number that sets how hard it is to find a valid block. The network adjusts it so blocks arrive roughly every 10 minutes no matter how many machines are hashing. More hashrate shows up, blocks come faster, and the next retarget raises difficulty to push the interval back to ten minutes.

After the 19.09.2026 retarget, difficulty stood at 132.76 T with the 7-day SMA of network hashrate at 943 EH/s (Hashrate Index, 14.09.2026, pre-retarget snapshot).

How does the 2016-block retarget work?

Nodes measure how long the last 2016 blocks took and compare it against the target of 20,160 minutes (2016 times 10). Faster than target means difficulty goes up. Slower means it comes down. The new value holds for the next 2016-block window, about two weeks.

Two limits people forget:

• A single retarget cannot move difficulty by more than a factor of 4 either way (a hard cap of +300% and -75%).
• The retarget lands on the window boundary, not on a calendar date. A slow network pushes the date later. A fast network pulls it earlier.

Typical adjustments in 2026 have been in the range of plus or minus 5-15%, per POOL BTC pool data.

How do I forecast the next difficulty adjustment myself?

You need two inputs: how many blocks the current epoch has produced, and how long they took.

1. Take the current block height and its remainder modulo 2016. That is your position in the epoch.
2. Work out the average interval: elapsed time since the first block of the epoch, divided by the blocks mined so far.
3. Apply the formula: change = 600 / average interval in seconds, minus 1.

Worked example. At an average interval of 570 seconds, 600 / 570 = 1.0526, so the expected adjustment is about +5.3%. At 625 seconds, 600 / 625 = 0.96, an expected drop of about 4%.

Accuracy improves as the epoch fills. Over the first couple of hundred blocks the spread is wide, since mining is a Poisson process and a run of ten fast blocks says nothing about the trend. From roughly the midpoint onward the estimate usually stays within a few tenths of a percent of the final figure. The 14.09.2026 forecast was +4.7% to +5.3%; the actual adjustment on 19.09.2026 came in at +4.16%.

The live estimate, without any arithmetic on your side, is on the difficulty forecast page.

How does difficulty turn into hashprice?

Hashprice is revenue per unit of hashrate per day, usually quoted in dollars per PH per day. It folds the block subsidy, the transaction fee share, the BTC price and network difficulty into one figure.

Difficulty sits in the denominator. Your share of the network equals your hashrate divided by the total. When the total grows, your share shrinks and hashprice falls even if BTC holds its price. It works the other way too: the monthly average hashprice low in June 2026 was $27.7 per PH per day (CoinShares Q2 2026 report, cited 16.09.2026), with roughly a fifth of the network estimated to be mining at a loss.

Practical takeaway: if your BTC income slides while your worker hashrate holds steady, check the last retarget date before you blame the pool.

Difficulty up X percent equals income down how much?

Not by X. Income is inversely proportional to difficulty, so the drop equals X / (1 + X). A 5% rise costs 4.76% of revenue; a 2% rise costs 1.96%.

The base in the table below uses the post-retarget hashprice of $40.93 per PH per day (21.09.2026). At a 4% pool fee, revenue per 100 TH/s per day: 100 * 40.93 / 1000 * 0.96 = $3.93.

BTC price is frozen here. In practice price swings harder than difficulty and can wipe out the whole table in a day. The difference is that difficulty moves predictably and mostly in one direction over long stretches, while price does not.

What does a difficulty increase do to net profit on a specific ASIC?

Far more than it does to revenue. Electricity is not indexed to difficulty: the network grows, the bill stays put, and the whole hit lands on your margin.

Take the Antminer S21 XP at 270 TH/s, 13.5 W/TH (Bitmain spec page, checked 23.09.2026). That is 87.5 kWh a day. At the post-retarget hashprice of $40.93/PH/day, gross revenue after 4% fee: 270 * 0.03929 = $10.61.

Revenue lost 4.76%, but at $0.08 net profit fell by about 14%. At $0.12 the machine flipped from barely positive to negative. The closer you sit to breakeven, the more each percent of difficulty hurts, and the sooner you need a curtailment plan.

Model your own case in the profitability calculator and compare machine efficiency in the miner rankings.

The margin leverage effect: why thin margins amplify difficulty

This deserves its own section because it is the single most misunderstood thing about mining economics.

If your machine grosses $10 and electricity costs $4, net is $6. A 5% difficulty cut to revenue is $0.50 off the top, leaving $5.50 net, a drop of 8.3%.

If your machine grosses $10 and electricity costs $9, net is $1. The same $0.50 cut leaves $0.50 net, a drop of 50%.

Same machine, same retarget, same 5%. The margin leverage is what makes high-tariff miners the first to shut down after a retarget, and it is why the shutdown tariff for a given W/TH shifts with every adjustment.

At the 19.09.2026 difficulty of 132.76 T and hashprice $40.93/PH/day, the breakeven efficiency at $0.08/kWh is about 20.4 W/TH. After a hypothetical next +5% retarget, it drops to roughly 19.6 W/TH. A machine sitting between those two numbers is profitable one week and unprofitable the next.

Current shutdown thresholds for different W/TH classes are in the shutdown tariff breakdown.

What is happening to hashrate and difficulty in September 2026?

Verified data points from Hashrate Index and CoinShares:

CoinShares expects up to 35 EH/s to leave the public mining group by the end of 2026, as operators convert sites to AI and HPC hosting (CoinShares Q2 2026 report). AI leasing currently pays about three times what BTC mining pays per megawatt ($1.5M vs $0.5M annual profit, CoinShares 17.09.2026). Keel (formerly Bitfarms) stopped mining entirely on 29 June 2026. Core Scientific paid $41.9M to cancel delivery of ~15 EH/s of next-gen chips and redirect to AI colocation.

If that hashrate does leave, difficulty will adjust downward, and the remaining miners will earn proportionally more per TH. That is the self-correcting mechanism: difficulty follows hashrate, and hashrate follows profit.

How to model income over multiple retargets

Do not multiply one day's net by 365. That number is fiction.

Instead:

1. Start with today's daily net (revenue minus electricity) for your hardware.
2. Pick a difficulty growth scenario. A conservative scenario: assume 8-12% monthly difficulty growth. At the current pace, income per TH/s decreases roughly 50% per year.
3. Break the year into 26 two-week windows. In each window, reduce revenue by the assumed retarget percentage. Electricity stays the same.
4. Sum all 26 windows. That is your modeled annual income.
5. Compare against the hardware cost plus any operating overhead.

This is still a model, not a guarantee. The actual path depends on BTC price, new hardware deployments, energy costs worldwide, and the AI hosting migration. But it is a far better model than "today times 365."

Run scenarios with different growth rates. At 2% per retarget, annual income is roughly 60% of what the flat projection says. At 4% per retarget, it is closer to 35%.

Where do I watch difficulty and hashprice?

The minimum useful dashboard covers four things: current difficulty, the estimated next adjustment, blocks remaining until the retarget, and hashprice in dollars per PH per day. All sit on the difficulty forecast page.

A sensible cadence: check once a week around the middle of the epoch, then once more a day before the retarget. Daily checks add little, since nothing inside the epoch changes except the precision of the estimate.

Can difficulty go down?

Yes. When total network hashrate falls, blocks arrive slower than the 10-minute target and the retarget goes negative. It happens after large sites go offline, during seasonal energy curtailment, or following a long hashprice slump that pushes inefficient hardware off the grid. The June 2026 difficulty drop is covered in Bitcoin difficulty drops 10%: what it means.

With the AI hosting migration pulling hashrate out of public mining fleets, downward adjustments may become more common. But the network has historically trended upward over any period longer than a quarter.

FAQ

How do I know when the next difficulty adjustment lands?

Take the blocks remaining in the epoch and the current average interval, multiply, and convert to hours. A ready countdown is on the difficulty forecast page.

How accurate is a mid-epoch adjustment forecast?

By the midpoint the estimate is usually close to final, typically within a fraction of a percent. Early in the epoch the spread is wide because block discovery is random. The 14.09.2026 forecast ranged from +4.7% to +5.3%; the actual result was +4.16%.

Can difficulty go down?

Yes. When total network hashrate falls, blocks arrive slower and the retarget goes negative. It happened in June 2026 when hashprice hit $27.7/PH/day and inefficient hardware went offline.

Why did my income drop more than difficulty rose?

Because your electricity bill did not move. A 5% difficulty rise takes 4.76% of revenue, and all of it comes out of margin. On an S21 XP at $0.08/kWh that is about 14% of net profit.

How often does Bitcoin difficulty adjust?

Every 2016 blocks, approximately two weeks at normal block times. The adjustment can be up to roughly plus or minus 300% per period but typically moves 5-15%.

Should I time my ASIC purchase around a difficulty drop?

The machine earns more per day after a drop, but difficulty drops are rare, short, and unpredictable. If the purchase makes sense at current difficulty, a drop is a bonus. If it only makes sense after a drop, you are betting on an event you cannot schedule.