Efficiency Class Reference

The five-tier efficiency classification used by System Cypresshub to compare mining equipment at 0.05 PH/s and above under Korean operating conditions.

Our efficiency classes rank mining hardware by measured joules per terahash (J/TH) after a 72-hour sustained bench run at ambient temperatures between 18°C and 28°C. Classes apply to equipment at 0.05 PH/s aggregate hash rate and above.

Elite Efficiency

Range: Below 24 J/TH measured

Units in Class A represent the current best practical efficiency for air-cooled ASIC deployments. At KEPCO industrial tariffs, these rigs deliver the lowest electricity cost per terahash. Examples from our reference database include recent-generation units with optimised firmware at stock settings.

Typical use: Long-term deployments where power cost dominates ROI calculations. Facilities with limited cooling capacity benefit from lower heat output per terahash.

Strong Efficiency

Range: 24–28 J/TH measured

Class B rigs offer a solid balance of acquisition cost and operating efficiency. Most new procurement in Gyeonggi facilities targets this tier when Class A units carry a significant price premium.

Typical use: Standard warehouse deployments with adequate ventilation. Suitable for operators expanding capacity without premium hardware budgets.

Moderate Efficiency

Range: 28–33 J/TH measured

Class C units remain viable where electricity costs are partially offset by low acquisition price — particularly second-hand lots. Margins narrow quickly if deployed at scale without tariff advantages.

Typical use: Short-to-medium term deployments, backup capacity, or sites with favourable power contracts. Monitor monthly costs closely.

Below Average

Range: 33–40 J/TH measured

Class D hardware struggles to remain profitable at standard Korean industrial rates unless hash price rises significantly or power is subsidised. We often find these units in mixed lots where sellers advertised a higher class based on manufacturer ratings rather than measured performance.

Typical use: Retirement candidates, spare parts sources, or temporary capacity during hardware lead times. Not recommended for new procurement.

Substandard

Range: Above 40 J/TH measured

Class E units fail our minimum threshold for cost-effective operation in South Korea under current tariff structures. Degraded hash boards, failing fans, or severely outdated firmware commonly cause this classification in used equipment.

Typical use: None for production mining. Units receiving Class E in a pre-acquisition audit should not be purchased for deployment.

How Classes Are Assigned

  1. Unit runs at manufacturer-recommended voltage for 72 hours
  2. Power draw logged at one-second intervals via calibrated meter
  3. Hash rate averaged from pool-reported shares over the same period
  4. J/TH calculated from stable-period averages (excluding first 6 hours of warm-up)
  5. Class assigned based on the ranges above

Important Notes

  • Classes reflect bench conditions. Poor facility cooling can shift real-world performance down by one class.
  • Firmware version is recorded with every assignment. Overclocking voids the class relative to stock settings.
  • Our classes are independent of manufacturer marketing tiers.

Request a Comparison

If you are choosing between models and need to know which class each will land in at your target hash rate, request an Efficiency Class Comparison Report or contact us with your hardware shortlist.