Thank you for the log. This is really unusual as in virtually all cases we have seen the GPU thread either crashes or freezes entirely. The watchdog checks for this but if there is any regular progress in the GPU thread, it "thinks" that the GPU is OK and doesn't restart the miner. In any case, there is something wrong with GPU2 (maybe the clocks are too high?) as it had 10 incorrect shares and 33 correct ones: almost one quarter of its shares are incorrect. Even if it is showing 31-32 MH/s hashrate, its effective hashrate is at most 80% of this - 25-26 MH/s. Generally, anything above 1-2% incorrect shares is a sign to dial back the overclock a little on that specific GPU to avoid losing too much of the effective hashrate.
We could add a command line option that specifies the minimal "normal" speed but it would be of limited usefulness if the rig has vastly different GPUs, or if the ambient temperature slowly rises and the GPUs start to throttle to avoid overheating.
Thank you for your fast reply,
Yes I'm having trouble with that Gpu since when I started mining, Thanks for your advice, I will try to clock it down..
It will be very helpful if to add a command like -minspeed in claymore, if it is also possible to set it per GPUs, instead of the total hashrate..
Ex:
If one of the GPUs hash rate didn't exceed 20MH/s for 5mins it will have an option to restart the miner or execute the reboot.bat
Thanks again! And more power!
Show Posts
