BTC updated

Ambient Temperature Derating for ASIC Miners

Ambient Temperature Derating for ASIC Miners

Every ASIC ships with a rated operating temperature range, and every ASIC also ships with firmware that throttles hashrate when chip or intake temperatures exceed safe thresholds. Operators who ignore the rated range either lose hashrate silently or damage chips outright. This article covers ambient temperature derating asic essentials: the manufacturer-stated operating ranges, the throttle behavior that kicks in above the limits, the climate-zone implications for siting decisions, and the practical mitigations operators apply at home and farm scale.

What “rated operating temperature” actually means

Manufacturer spec sheets typically list two temperature numbers. The first is the rated operating ambient — the intake-air temperature the unit can run at full hashrate indefinitely. For modern Bitmain Antminer S21 family units, this is typically 5°C to 35°C (41°F to 95°F). For Microbt M-series, similar. Canaan Avalon Q-series and A-series publish comparable ranges. The second number is the storage temperature range, which is wider and doesn’t affect operating decisions.

The “full hashrate indefinitely” caveat matters. Inside the rated range, the ASIC delivers the spec-sheet hashrate without thermal throttling. Above the rated range, firmware progressively reduces hashrate to keep chip temperatures within safe bounds. Below the rated range, units sometimes have trouble starting reliably — extreme cold causes its own set of problems, especially condensation when the unit warms up. The Bitmain Antminer catalog and the Canaan Avalon catalog publish per-model temperature ranges.

How firmware throttles above the limit

Modern ASIC firmware monitors per-chip temperatures (typically 60–90 sensors per unit across the hashboards) and reduces clock speed when any chip exceeds its safe threshold (typically 90–100°C junction temperature for current-generation chips). The throttle is gradual: a few percent hashrate reduction at first, climbing to 20–40 percent reduction as ambient temperatures push 5–10°C above the rated maximum. Above that, units may shut down entirely to prevent damage.

The user-visible signature of thermal throttling is wattage staying constant (or rising slightly as fans speed up) while hashrate drops. Operators with per-circuit power monitoring from the power meter monitoring guide see this clearly: J/TH efficiency degrades because the ASIC is doing less work per watt. The hashrate dropout troubleshooting guide covers the diagnostic workflow when the cause isn’t obviously thermal.

Climate-zone implications for siting

Operators selecting sites for farms or home installations need to think about local design ambient temperatures, not annual averages. A site that averages 15°C annually may hit 35°C+ during summer afternoons, and the cooling system has to handle the peak, not the average. The U.S. EIA and ASHRAE publish 99 percent design temperatures by metropolitan area — the temperature exceeded only 1 percent of hours in a typical year — which is the right input for cooling system sizing.

Cold climates dominate the conversation for a reason. Sites in North Dakota, Montana, Washington, and similar regions rarely exceed 30°C and often operate at sub-zero intakes for half the year. Air-cooled deployments in these climates rarely face thermal throttling and avoid the capital cost of evaporative or mechanical cooling. The tradeoff is winter operations: condensation when warming up after extended cold soaks, and the need to keep units running continuously to avoid cold restarts.

Hot-climate mitigation strategies

For sites in Tennessee, Georgia, Oklahoma, and similar regions where summer ambient regularly exceeds 35°C, thermal mitigation becomes a recurring planning problem. Three approaches dominate. First, evaporative cooling: pulling outside air through a wet media that drops temperature 5–10°C in dry climates and 2–5°C in humid ones. Capital cost is modest and operating cost is low; the limitation is humidity.

Second, mechanical cooling: full HVAC with chillers or DX units. Capital and operating cost are significant but full control over intake temperature is available. Third, shifting to hydro or immersion cooling: covered in the hydro vs air cooling guide and the immersion vs hydro cooling tradeoffs. Liquid cooling tolerates higher ambient temperatures than air cooling because the heat-rejection equipment can operate at higher temperatures than ASIC chips themselves.

Time-of-day and seasonal patterns

Even in temperate climates, ambient temperature varies dramatically across a 24-hour cycle. Sites that operate near the throttle threshold during summer afternoons can run at full hashrate during cool nights and early mornings — a daily cycle of partial throttling. Operators can sometimes shift maintenance windows to the hottest hours when hashrate is already degraded, minimizing the marginal cost of downtime.

Seasonal patterns are similar. A site that comfortably runs air-cooled from October through April may need supplemental cooling from June through September. Operators in transitional climates often design for the worst summer week and accept that the infrastructure is overbuilt for 9 months of the year. The capital cost of the cooling system has to be amortized against full-year operations even if it’s only fully utilized for 100 hours per year.

Intake temperature vs ambient: the difference that matters

Manufacturer rated ranges are for intake temperature, not for outdoor ambient. The difference is the temperature rise between outdoor air and the ASIC’s actual intake — driven by short-circuit recirculation, building thermal mass, and any pre-cooling or pre-heating in the air path. A facility with outdoor ambient at 30°C may deliver intake at 32°C (well-managed airflow), 38°C (poor aisle layout with mixing), or 25°C (with evaporative pre-cooling).

Operators should monitor intake temperature at each unit (firmware reports this on the dashboard) rather than relying on outdoor weather data. Persistent intake temperatures above ambient indicate airflow problems that need attention — usually short-circuit recirculation that the hot/cold aisle layout is designed to prevent. The ducting and airflow design guide covers the airflow patterns that hold intake-to-ambient differential under 3°C.

Voltage and overclock interactions with thermal limits

Operators running firmware overclocks (Vnish, BraiinsOS, LuxOS) push chip temperatures higher at the same ambient, narrowing the thermal headroom. A unit that runs full hashrate at 32°C ambient on stock firmware may throttle at 28°C ambient under aggressive overclock. Operators planning overclocks should derate the effective operating ambient by 3–8°C from the manufacturer’s stock rating.

Conversely, undervolting (reducing chip voltage at stock or slightly reduced hashrate) increases thermal headroom and lets units run at higher ambient without throttling. The ASIC undervolt safe ranges write-up covers the voltage-vs-temperature tradeoff. Operators in hot climates often undervolt by 5–10 percent and accept slightly lower hashrate to maintain stable operation through the summer.

Cold-weather considerations

Cold isn’t a free ride. Below 5°C ambient, several issues appear. ASIC PSUs can have trouble starting in extreme cold (some units rate down to -5°C, others not below 0°C). Condensation forms on internal components when units warm up from cold soak — the moment power applies and chips heat from -10°C ambient to operating temperature, atmospheric moisture condenses on the warming surfaces. Operators in cold climates often keep buildings minimally heated above freezing and rely on continuous operation to prevent cold-soak conditions.

Heat recovery becomes attractive precisely in these climates — turning the mining waste heat into building heat is essentially free thermal management. The mining as heat source overview and the home garage mining guide cover the integration patterns.

Practical operating range for planning

For planning purposes, operators should design air-cooled mining sites to keep intake temperatures between 10°C and 30°C across the operating year. That gives 5°C of headroom against the typical 35°C upper rated limit, which is enough to absorb the 3–5°C of typical intake-to-ambient rise plus occasional unusual conditions. Sites designed to this spec rarely face thermal throttling.

Sites that allow intake temperatures to push 35°C+ should expect 5–15 percent hashrate degradation during peak periods, which compounds annually into meaningful revenue loss. Operators evaluating bitcoin mining profitability in hot climates need to model this throttling into their hashrate assumptions, not assume nameplate hashrate year-round.

Selecting hardware to match the climate

Operators in hot climates should prioritize models with the highest rated operating temperatures and the best thermal headroom. The Coin Web Mining catalog publishes operating temperature ranges on every product page, and hydro/immersion variants of the same generation typically offer significantly higher effective operating ambients than air-cooled siblings — the liquid loop carries heat away regardless of room ambient.

For air-cooled deployments in hot climates, evaporative or mechanical pre-cooling of intake air is often more cost-effective than upgrading to hydro at small scale. The breakeven shifts toward hydro/immersion as unit count grows past 20–50 units, at which point the shared infrastructure costs amortize favorably.

For operators evaluating a new site in an unfamiliar climate, the right diligence is to pull at least three years of hourly weather data for the location and model expected derating against the manufacturer’s published throttle curves. A site that looks fine on annual-average temperature can hide a hundred hours per year above the throttle threshold, which translates directly into lost hashrate and shifted ROI math. Most modern climate datasets (NREL, NOAA) are free; the analysis takes hours, not days.

References

What's the maximum operating temperature for an Antminer S21?
Bitmain rates the S21 family for 5°C to 35°C intake-air temperature for full hashrate operation. Above that range, firmware progressively throttles hashrate to keep chip temperatures safe. Exact figures vary by model and revision — pull the spec sheet for the specific unit.
How much hashrate do I lose at 40°C ambient?
Typically 10–25 percent depending on the model and firmware. Throttle is gradual — small reductions at 36°C climbing to significant cuts by 40°C. Above 42–45°C ambient, units may shut down entirely to prevent damage.
Should I undervolt in hot climates?
Often yes. Undervolting by 5–10 percent reduces chip heat output and increases thermal headroom, allowing stable operation at higher ambient temperatures. The tradeoff is slightly reduced hashrate, but the alternative is sustained throttling which loses more hashrate net.

The miners in the Coin Web Mining catalog publish operating temperature ranges on every product page, helping operators match hardware to the local climate. Bulk farm orders can request a quote for five units or more.