Releases: develsoftware/GMinerBetaRelease
Releases · develsoftware/GMinerBetaRelease
v2.86
2.67
- partial LHR cards unlock without additional power consumption
- use --lhr 1 to force enable unlock for RTX 3060 V1
- use --lhr_tune1/lhr_tune2 to tune unlock parameters, positive values increase performance, negative values decrease probability of lock
2.55
- lock core clock feature for Nvidia GPUs under Windows and Linux (--lock_cclock)
- resolve domain names even with broken system DNS settings
- support SOCKS5 proxy (--proxy host:port)
2.51
- significant CPU usage reduction (up to 5x, fixed LA problem under HiveOS)
- fixed bug with AE mining (appeared in 2.50)
- fixed bug with worker name (appeared in 2.50)
2.50
- auto fan feature, now you can set target fan temperature (--tfan, Windows only)
- reset overclocking before DAG generation (Windows only)
- improved beamhash performance (~+2%)
- improved cuckatoo32 performance (~+5%)
- added display of GPU model in statistics table
- added --worker parameter to specify worker name for ETH pools thats doesn't support wallet.worker notation
- added option to display date in log (--log_date)
- added option to log stratum (--log_stratum)
- display epoch and block number on new job
- updated Web UI for --api
- uses --proto stratum for NiceHash by default
- bug fixes and stability improvements
2.48
Windows only version (there is no changed for Linux)
- minimized memory leaks cause by Nvidia NVML library under Windows
- fixed memory tweaks support for Mining Edition GPUs under Windows
- added options to control overclocking under Windows (fan speed, power limit, core clock, core voltage, memory clock, memory voltage. check readme.txt for more details)
UPD Beta2:
- fixed bug with --ssl option
1.89
- added solo mining support for kadena
1.85
- added blake2s algorithm for Nvidia GPUs (Kadena coin), hope dual mining with ETH will be added in next update
- decreased stale shares percentage on pool side for cuckatoo31 algorithm
1.80
- added Cortex algorithm for Nvidia GPUs
Minimum GPU memory requirements is 8GB
DevFee on Cortex algorithm is 5%
1.74
- significant performance improvements for Cuckatoo31 algorithm