- 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-platformis 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
- Optimizations to whirlpool (+65%), bmw (+50%), and echo (+3%) kernels for an overall 6% speedup in x16.
- Significant optimizations to Hamsi (+35%). +4% overall increase in average x16r/x16s hashrate.
- 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.
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.