18 August 2025 · 5 min read
A mismatched power supply is one of the most common reasons ASIC units fail our intake inspection before testing even begins. The unit may hash correctly on a bench PSU, but voltage sag under load produces unstable readings and can damage hash boards over time.
Voltage and Phase Requirements
South Korean commercial facilities typically provide 220V single-phase or 380V three-phase depending on the building classification. Most ASIC PSUs shipped to this market accept 200–240V input. Verify the input rating label before connecting — units configured for 110V-only input will not survive Korean mains voltage.
Continue reading →1 August 2025 · 7 min read
South Korean summers push ambient warehouse temperatures past 30°C in facilities without dedicated HVAC. Operators who chase maximum hash rate through aggressive overclocking often see unit failure rates climb within 18 months. Our assessment practice deliberately uses conservative cooling profiles because most of our clients plan for three-to-five-year hardware lifecycles.
What Conservative Means on Our Bench
- Ambient held at 24°C (not the 18°C some manufacturers use)
- Stock firmware, no overclock profiles
- Fan speed left on automatic — we do not force low-RPM “quiet modes” that raise chip temperatures
- Underclock test included: a single step down to document the watts-per-terahash trade-off
The Data Point Operators Request Most
Watts per terahash at stock settings under our cooling profile. This single figure helps compare models with different nominal hash rates on equal footing. An S19j Pro at 29.4 J/TH under our protocol often outperforms an older S17 Pro at 39.4 J/TH, even though both clear the 0.05 TH/s floor comfortably.
Continue reading →3 July 2025 · 8 min read
Manufacturer datasheets list power consumption under ideal laboratory conditions: 25°C ambient, fresh fans, balanced three-phase input. A warehouse in Hwaseong in August looks nothing like that laboratory.
Our Measurement Setup
We connect the ASIC’s power supply directly to a Yokogawa WT310E through a calibrated PDU. Readings are logged every 15 seconds over a minimum four-hour window. Ambient temperature is held at 24°C with intake airflow at 1.2 m/s — deliberately conservative, cooler than most operating environments but warmer than a cold-storage test chamber.
Continue reading →12 June 2025 · 6 min read
When operators ask why we decline to assess USB-stick miners or legacy Block Erupters, the answer comes down to measurement resolution. Our bench instruments — a Yokogawa WT310E power analyser and a calibrated hash rate counter — produce reliable data only when the unit under test draws enough current to exceed the analyser’s noise floor.
At hash rates below roughly 0.05 TH/s (50 GH/s), signal-to-noise ratio in wattage readings becomes unreliable. A device drawing 15 watts can appear to fluctuate by ±3 watts depending on USB hub quality, cable length, and ambient EMI. That margin of error makes any report we issued legally indefensible.
Continue reading →