RGminer v0.9.8

6 juil. 2026

rgminer 0.9.8

This release focuses on improved Pearl performance and energy efficiency across NVIDIA Turing, Ampere, Ada and Blackwell GPUs.

Highlights

  • Ampere: around +1% hashrate with about -2% power draw
  • Ada / Blackwell: up to +11% hashrate
  • Turing: new best energy-efficient kernel profile

Performance examples

GPU Architecture OC settings Hashrate Power Efficiency Power per TH
RTX 2080 Turing 125 / 1650 / 5000 43.66 TH/s 160 W 0.273 TH/W 3.66 W/TH
RTX 3070 Ampere 250 / 1450 / 5000 65.5 TH/s 133 W 0.492 TH/W 2.03 W/TH
RTX 4070 Ti Ada 350 / 2450 / 5000 135.6 TH/s 188 W 0.721 TH/W 1.39 W/TH
RTX 5070 Ti Blackwell 490 / 2450 / 5000 153 TH/s 205 W 0.746 TH/W 1.34 W/TH

Backend policy

  • Turing, Ampere and Ada builds are shipped on CUDA 12.4
  • Blackwell build is shipped on CUDA 12.8

Performance depends on GPU model, driver, clocks, memory and power limit conditions.