Skip to content
Compare
Choose a tag to compare
  • Optimizations to Echo (+7%) for an overall 1% speedup in x16.
  • Added code to automatically select AMD OpenCL platform by default. This means that setting --gpu-platform is no longer required for rigs with integrated graphics or Nvidia cards.
  • More sane default parameters. Changed default xIntensity to 256, from a default intensity of 8.
Compare
Choose a tag to compare
  • Optimizations to whirlpool (+65%), bmw (+50%), and echo (+3%) kernels for an overall 6% speedup in x16.
Compare
Choose a tag to compare
  • Significant optimizations to Skein (+80%), Blake, (+65%), and Keccak (+40%), for an overall 3.5-4% increase in hashrate.
  • Dev pool switching is now much less disruptive (no longer reinitializes GPU threads). This should fix issues like these.
Compare
Choose a tag to compare
  • Significant optimizations to Hamsi (+35%). +4% overall increase in average x16r/x16s hashrate.
Compare
Choose a tag to compare
  • Added initial xevan algo support (only seems to work with RX-series cards).
  • Added warning that building kernels may take a few minutes the first time.
Compare
Choose a tag to compare

Avermore initial release. Initial tests show 10% speedup over ace's sgminer-gm-x16r.

Optimized x16r / x16s kernels.

1% devfee (1 minute every 100 mins) to support continued work on this miner.