Downloads

Benchmarks

Reproducible baselines for desktop CPU & GPU miners

Target hashrate, power and efficiency per coin and miner — grounded in runs you can repeat with the same versions, clocks and pools.

View reference rigs Go to docs
Reference rigs Coin matrix Method Usage notes

Reference rigs

Baseline runs you can copy

Each card lists miner version, clocks and power so you can reproduce the numbers on your own rigs.

RTX 4070 Ti · Ravencoin

NBMiner 0.9x · PL 235 W · core -200 · mem +1200 · driver 536.xx

  • Hashrate: ~47 MH/s KawPoW
  • Wall power: ~245 W
  • Efficiency: ~0.19 MH/J

RX 6800 XT · Ergo

TeamRed 0.10.x · core 1150 · mem 2150 · PL 150 W · driver 23.x

  • Hashrate: ~165 MH/s
  • Wall power: ~175 W
  • Efficiency: ~0.94 MH/J

Ryzen 9 7950X · Monero

XMRig 6.x · 16 threads · huge pages on · SMT off · 3600 MT/s RAM

  • Hashrate: ~45 kH/s
  • Wall power: ~180 W
  • Efficiency: ~0.25 kH/J

RTX 3060 12GB · Kaspa

lolMiner 1.8x · core 1500 · mem -1000 · PL 115 W · driver 531.xx

  • Hashrate: ~55 MH/s
  • Wall power: ~120 W
  • Efficiency: ~0.46 MH/J

Per-coin matrix

Pick a coin and see typical targets

Use these to sanity-check new rigs or spot when your miner drifts. Adjust for your own power price.

Ravencoin · GPU

Best on NVIDIA; memory helps less than core stability.

  • 4070 Ti: ~47 MH/s @ 245 W (NBMiner 0.9x)
  • 3060 12GB: ~23 MH/s @ 125 W
  • Alert if stale shares > 3%.

Ergo · GPU

Memory-bound; AMD often wins on efficiency.

  • 6800 XT: ~165 MH/s @ 175 W (TRM 0.10.x)
  • 3060 Ti: ~115 MH/s @ 125 W
  • Watch for VRAM errors when pushing mem.

Kaspa · GPU

Core-heavy; responds well to lower PL with balanced clocks.

  • 3060 12GB: ~55 MH/s @ 120 W (lolMiner 1.8x)
  • 3070: ~66 MH/s @ 125 W
  • Keep temps below 70°C for stability.

Monero · CPU

Depends on memory latency and huge pages; avoid SMT thrash.

  • 7950X: ~45 kH/s @ 180 W (XMRig 6.x)
  • 5900X: ~26 kH/s @ 140 W
  • Alert on hash drop after 15 min warmup.

Method

How we record a run

Repeatable steps keep benchmarks honest. Use the same flow when sharing your numbers.

Test recipe

  • Fresh driver install, reboot, no background renders.
  • 20–30 minute warmup before logging hashrate.
  • Record ambient temp, clocks, PL/voltage, fan curve.
  • Use the same pool endpoint for comparability.

What to compare

  • Hashrate stability over a 15 minute window.
  • Pool-side vs miner-side accepted shares.
  • Net $/kWh when toggling efficiency vs performance.
  • Noise/thermals vs your location limits.

Usage notes

Make benchmarks work for you

Keep rigs predictable and avoid chasing short peaks that cost more power than they return.

Track deltas

Keep a log per rig with miner version, clocks, pool and wall draw. Flag runs where pool-side rewards diverge from baseline.

Tune for your power price

High $/kWh: down-volt and chase efficiency. Cheap power: test performance mode but watch rejects.

Share safe configs

When a rig is stable, save the profile and reuse it across similar GPUs/CPUs. Note exact driver and miner build so teammates can reproduce.