Quick answer
This is D-Central's open bench-measurement dataset: real wall-power, hashrate, J/TH efficiency and noise (dBA) figures for ASIC miners, measured first-hand on our Laval bench rather than copied from spec sheets. We add one model per month and publish exactly how each figure is taken. The dataset is live and the schema is fixed; the first measured entries are pending the opening bench session.
Free to download (CSV + JSON, CC BY 4.0). Every figure links to the published method, so you can reproduce or challenge it.
What this dataset records
Spec sheets quote idealized numbers; the wall and the sound-meter do not lie. For each miner and tuning mode we publish a fixed set of fields, measured the same way every time:
| Field | What it is |
|---|---|
| Miner model | The exact miner model and hardware revision under test. |
| Tuning mode | The tuning mode (stock, eco/undervolt, balanced, or turbo/overclock). |
| Wall power (W) | Power drawn at the wall outlet — what your meter bills you, including PSU losses. |
| Hashrate (TH/s) | Sustained hashrate at that mode, read from the miner after warm-up. |
| Efficiency (J/TH) | Efficiency, computed as wall watts ÷ TH/s. Always derived from the two raw figures so it can never drift. |
| Noise (dBA @ 1 m) | Sound pressure in A-weighted decibels, measured at 1 metre — the number that decides whether it can live in your home. |
| Measured (date) | When the measurement was taken (the network and ambient conditions move, so the date matters). |
| Method | A link to the published methodology so every figure is reproducible. |
Measurements
First entries pending the opening bench session.
We are setting this dataset up the honest way: the schema, the download endpoints and the methodology are live now, and the measured rows go in as they are taken — we are not back-filling it with manufacturer spec-sheet numbers dressed up as bench data. The opening session is being scheduled; the first measured model will appear here, then a new model every month.
Download CSV Download JSON CC BY 4.0
One model a month, measured the same way every time
This is a standing program, not a one-off: one model is benched and published each month, building a slow, trustworthy library of real-world efficiency and noise figures. Because bench measurements are physical, accuracy beats speed — each entry is taken to the same published methodology (wall power at the plug, hashrate after warm-up, J/TH computed, noise at 1 metre). We bench the open-source hardware we build and the miners we repair; we do not publish numbers we did not measure.
Cite this dataset
This dataset is free to use, share and adapt under the Creative Commons Attribution 4.0 International (CC BY 4.0) licence — please credit D-Central and link back.
Suggested attribution:
D-Central Technologies. "D-Central Bench Measurements (efficiency + noise)." d-central.tech, updated 2026-06-13. Licensed under CC BY 4.0. Accessed 13 June 2026. https://d-central.tech/data/bench-measurements/Related products, repair, and setup paths
- how D-Central diagnoses ASIC repairs
- ASIC troubleshooting library
- ASIC manuals and repair guides
- replacement hashboards
- ASIC control boards
- ASIC power supplies
- S19 family replacement hashboard
- C52 replacement control board
- APW12 S19 power supply
- immersion cooling hub
- home immersion cooling guide
- ASIC miners for immersion planning
- ASIC cooling parts
- airflow shroud before immersion
- compare miner specs in the database
- ASIC repair support
Last reviewed June 12, 2026.