How much energy does a post-quantum chain burn?
Quantus secures its blocks with real work — a Poseidon2 proof-of-work running on GPUs, not a validator list. So the chain has a genuine electricity footprint, and this desk measures it honestly: the network hashrate is implied from live difficulty, but the watts behind it depend on the fleet's efficiency, which nobody publishes. We refuse to print a fake single number. Everything below is a band, bounded by measured hardware on one side and a conservative worst case on the other.
The power band why a band, not a number
The chain publishes difficulty, not wattage. One fact — H = D / 12 — gives the hashrate exactly. Turning that into watts requires knowing how many joules each hash costs, and that is a fleet assumption. Drag the slider to see how the whole footprint scales with it. The two pins mark the measured fleet bounds.
Fleet mixer compose the network you believe in
A fleet's efficiency is the hashrate-weighted harmonic mean of its parts — slow cards drag the average down harder than they look. Mix measured GPUs and watch the band contract toward the fleet you actually expect to be out there.
The network power here rescales the live hashrate through your fleet's efficiency — a thought experiment, not a census.
Measured hardware the benches that bound the band
Every row below is a real measurement from the upstream miner team's bench docs — rented Vast.ai / Clore.ai GPUs running quantus-miner benchmark. Green badges are measured nvidia-smi power caps; amber badges are board TDPs, which overstate wall power and therefore understate efficiency.
| GPU | Engine | Hashrate | Power | Efficiency | Date | Source |
|---|
Rig builder your card, the live chain, honest math
Expected daily QTC from your rig's share of the live network hashrate, at the current emission-model reward. QTC has no market price yet, so instead of fake profits we show the break-even QTC price: what a coin would need to be worth for your electricity to pay for itself.
Assumes the reward and difficulty of the latest snapshot, and 24/7 mining at the preset's bench power. Card purchase, rent, downtime, fees, and luck are not in this number.
Energy history the footprint since genesis
Daily network energy, recomputed from every ~100th block's difficulty since genesis (1,450 samples), drawn at both efficiency bounds. The rise you see is the hashrate finding its level, not an accounting artifact.
—
Per-transfer energy the cost of a young chain
The network burns energy to make blocks, not transfers. Divide block energy by actual throughput and each transfer looks wildly expensive — because the chain is young and mostly empty. At its design throughput, the same power spreads over thousands of times more transfers.
Neither bar is a "cost per payment" claim — blocks are secured whether transfers ride along or not. The comparison shows how per-transfer energy collapses as throughput fills the chain.
Scale check one young chain against the industry
Annual electricity, log scale. The Quantus band is the whole-network estimate from the band lab above; the Bitcoin figures are the current Cambridge and Digiconomist estimates of the same network — they disagree, honestly, so we show both. Nothing here is adjusted for what the energy does.
Carbon desk intensity is the whole game
Energy is joules; emissions are joules × the grid they're drawn from. Set the grid intensity your fleet actually runs on and read the network's annual footprint — and how many U.S. homes it equals.
Home equivalence compares electricity only — it says "this much power", not "this much impact". The chain publishes no fuel-mix data; intensity is your assumption to make and to label.
Methodology & honesty read this before quoting
What is exact
H = D / 12: expected hashes to win a block equal difficulty D (verified in the Consensus Lab againstpallet_qpow), and the network targets one block per 12 s — so the implied hashrate is exact up to difficulty noise.- Efficiency arithmetic:
W = (H/s) / (H/J), annual energy =MW × 8760 h, per-block energy =W × 12 s. All inenergy-core.js, 97/97 node tests green. - Fleet efficiency is the hashrate-weighted harmonic mean — the only way the joules add up when cards differ.
- Reward math:
R = (21M − S) / 50M, exact to the planck against the emission source (see Emission Lab).
What is assumed
- Everything about watts. The fleet mix is unknown; the band edges are an RTX 4090 at a measured 350 W cap and an RTX 3060 Ti at board TDP. If real fleets run cheaper cards or undervolt, the truth sits higher or lower — say so when you cite this page.
- TDP badges are not measurements: wall power for those cards will differ, usually lower.
- Snapshots age: consensus and supply snapshots refresh with the loop's data scripts; the pill shows exactly how old they are. When both snapshot and a live read fail, the UI says so instead of inventing numbers.
What we did not do
- No fuel-mix claims for Quantus mining — there is no public survey of where the hashrate sits, and guessing would be fiction.
- No "greener than Bitcoin" claims — the chains differ by orders of magnitude in hashrate, security assumptions, and signature weight; per-TWh comparisons measure scale, not virtue.
- No unverified external figures: entries that couldn't be re-verified on Oct 1, 2026 were omitted rather than carried over.
Honest caveats
- Energy is consumed to secure blocks; per-transfer figures are arithmetic, not causation.
- QTC has no market price yet: revenue, profit, and payback figures cannot exist and are not shown.
- External figures are cited with source, URL, and date — check them before republishing; they move.
- This page never fabricates chain state: if the snapshot and the indexer are both unreachable, it says so.