Stock Antminer S9 research interface
The Live Lab studies the boundary between PC-side orchestration, mining jobs and a stock ASIC. Telemetry integration is pending. The nominal device capability is approximately 13.5 TH/s; no improvement over stock operation is claimed.
- Published
- 2026-08-22
- Updated
- 2026-08-22
- Authors
- BTC PoW Lab
- Replication
- Lab integration in progress
Local structure, statistical signal and operational advantage are evaluated separately.
Question
Can research strategies be tested through a stock-miner-compatible job interface without modifying ASIC silicon?
Hypothesis
Selected orchestration and candidate-scheduling experiments can be evaluated against real ASIC behavior.
Why this matters
Simulation-only savings can disappear when translated into protocols, firmware constraints and energy use.
Method
Prepare controlled jobs, preserve stock-miner compatibility, collect shares and observations, then compare against an equivalent baseline.
Experimental setup
- Baseline
- Unmodified stock S9 job processing.
- Setup
- PC intelligence ↔ mining job ↔ stock Antminer S9.
- Hardware
- Bitmain Antminer S9; nominal capability ≈13.5 TH/s.
- Dataset
- Telemetry integration pending; no live numbers are currently published.
Results
- Real hardware is available as a validation target.
- Telemetry publication is not yet connected.
- No hashrate improvement is claimed.
Scientific analysis
ASIC-S9-01 defines the measurement boundary between PC-side job construction and a stock Antminer S9. The purpose is to prevent software-only estimates from being presented as hardware results before protocol, firmware, thermal and electrical costs are observed.
The approximately 13.5 TH/s figure is a nominal device capability, not live telemetry. A valid campaign needs synchronized job IDs, accepted/rejected share counters, wall-clock windows, power measurements and environmental covariates.
Comparison will use equivalent jobs and repeated baseline/treatment windows. Hashrate alone is insufficient: stale work, rejected shares, power and thermal throttling can reverse an apparent gain.
Mathematical formulation
Ĥ = hashes or difficulty-equivalent work / elapsed timeEstimated effective hashrate must use a declared work estimator and measurement window.
η = P / Ĥ [J/TH]Energy efficiency combines measured electrical power P with effective hashrate.
Δ = metricstrategy − metricbaselinePaired or blocked windows reduce confounding from temperature, pool variance and job changes.
Statistical reading
Share counts are stochastic; short windows have high variance and cannot support precise speed claims.
Baseline and treatment should alternate or randomize across comparable thermal states, with confidence intervals reported for the paired difference.
Until telemetry is connected, the project reports integration status rather than simulated live numbers.
Validity and scope
- No hashrate-improvement claim currently exists.
- Nominal hardware specifications are not measurements of this unit.
- External validity will depend on firmware, frequency, voltage, cooling and job protocol.
Methodological references
Interpretation
The S9 is an experimental instrument, not a cloud-mining product or active public pool.
Limitations
- Nominal capability is not a live measurement.
- No public telemetry source is connected yet.
Reproduction
Future campaigns will publish job definitions, baseline conditions and measurement windows when safe.
Artifacts
Hardware interface documentation planned.
Experimental content is provided for research and educational use, without warranty. Validate independently before relying on it. Read the full disclaimer.