GoMining, computed.
data as of 2026-08-30Same money, same one-year horizon, all fees deducted: what is left at the end on each side — and difficulty keeps rising?
One calculation, once, on both sides of the same dollars: the money either buys a share of a real mining operation (how that works) or plain Bitcoin. The result, after every documented fee, is the table above, per price path. Editable fields are yours; pre-filled ones carry their source and date. The model behind it was checked day-by-day against a real account (how, and how well) — no sign-up, no promised number anywhere.
At the defaults ($1,000, 100 TH, no discount, +9% reinvestment growth, difficulty +20%), holding ends ahead in all three paths — by ≈$669 flat, ≈$711 if Bitcoin rises 50%, ≈$571 if it falls a third. At documented rates the miner yields about $1.21 per day per $1,000 invested — simple payback ≈27 months at today's rate (not a forecast; difficulty growth extends it). Holding starts free of any fee bill.
The observed levers close roughly half that gap: the reference account's 28.26% maintenance discount lifts the same setup to about $1.97 per day (gap ≈$378 flat) — and an honest exit price does the rest, because the tokenized mining power currently has no fixed expiry under the stated terms, though its BTC output declines as network difficulty rises (one market check: cheapest listing ≈96% of primary, seller ≈91% net of the 5% fee — today's observation, not a law). This page keeps exit value at zero; the Calculator carries it as an explicit lever with its payback math at today's rate (≈92% of purchase price after 12 months to tie holding).
Conditions, not predictions: mining wins when discounted, reinvested yield plus a defensible resale value outruns the coins the same dollars could have bought; holding wins whenever that fails. The table above shows both outcomes at once, per path.
Everything above assumes network difficulty grows +20% per year. That is a choice, not a fact — and the real world has been gentler lately. Measured on a real account's four-year daily series with the two halvings excluded, net difficulty grew +95% in 2023, +43% in 2024, +36% in 2025 — and in the twelve months to August 2026 only +10%, with 2026 so-far running negative (−11%) as weak hash-price pushed marginal hashrate offline. Difficulty is a thermostat, not a staircase: when harvest falls below the electricity cost of industrial miners, they switch off and it falls.
The same engine, same inputs, only the growth rate changed — cumulative mining cash on the newcomer setup ($1,000 → 100.1 TH, 15 W/TH, 0% discount, flat BTC price, no resale):
| Difficulty growth | Margin dies at | Cash after 1y | 3y | 10y | Real-world anchor |
|---|---|---|---|---|---|
| +20%/yr (site model) | ~2.0 years | $317 | $439 | $438 | 2024–2025 era |
| +10%/yr | ~3.9 years | $376 | $793 | $876 | last 12 months |
| +5%/yr | ~7.6 years | $409 | $1,038 | $1,752 | slow-growth years |
| ±0%/yr | never | $443 | $1,330 | $4,433 | 2022, 2026 (falling) |
Read it as: the +20% assumption makes every horizon on this site the early-death case. At the actual growth of the last twelve months, the same newcomer farm collects roughly twice as much before the margin closes; in flat-difficulty years it never stops earning. What stays true everywhere: without a maintenance discount the slow decay still never recovers the $1,000 — rate sensitivity changes how long the machine lives, the discount decides whether it was worth buying.
The model in this table has the same answer the network will have: the block reward drops from 3.125 to 1.5625 BTC, and every miner's gross line falls by half overnight. Nothing about the difficulty, your W/TH class or your maintenance bill changes that day. On this site's canonical newcomer setup ($1,000 → 100.1 TH, 15 W/TH), that pushes the daily gross from ≈$3.91 to ≈$1.96: a 0%-discount account turns negative and stops paying entirely, while a maximum-discount account (28.26%) keeps a knife-edge margin of ≈+2% per day — income roughly equal to its electricity and service bill.
Three levers decide which side of that line you sit on: the maintenance discount (the difference between +2% and −27% is the discount stack, not luck), the efficiency tier (the treadmill — 12 W/TH holds 2.08× margin where 50 W/TH is already underwater), and auto-growth (reinvested TH offsets part of the difficulty drift). The 24-month horizon on this page spans the halving; the engines flag it and let you set the block reward yourself instead of silently halving your result — see the Method page's deliberate model limits.
This engine was validated day-by-day against a real mining account over a full year (2025-08-30 to 2026-08-29): 329 of 357 days within ±2%, 346/357 within ±5%, mean deviation +0.03% (account data via the official GoMining API, 357 days retrieved 2026-08-30; see the Validation Log page and the Method page).
NetPerHash (domain registered 2026-08-28) is an independent research project, funded partly through one disclosed GoMining referral link: tokenized mining, computed instead of marketed. The figures are either public with source and retrieval date, or come from the operator's own account, published figure-by-figure with per-figure approval — no wallet addresses, deposit addresses or balances anywhere on this site.
Update log.
[Further entries appended here with dates as sources are re-verified monthly and figures move.]
All numbers computed in your browser from one calculation engine verified by automated tests. Sources retrieved 2026-08-30 unless marked live.