NBMiner
NBMiner copied to clipboard
NebuTech
CUDA error: an illegal memory access was encountered (err_no=77)
Tried many times, still get this same issue.
I used 42.2 windows version, 512.15 driver version.
I tried to downgrade both cclcok and mclock, but both failed... still don't understand why.
New issue for me today everything was fine for months, hashrates declined overnight and gpus reset to default settings.
The fix was drastically reducing mclock, I was at 1300/1100 and settled on 800mhz. My power limit stayed the same around 60% and cclock is still locked at 1380. Hashrate obviously decreased around 3-5mh but efficiency is around the same so nothing too dramatic. (These are my personal cards may not be the same for yours)
Good luck!
Notes: {NBminer 42.2} {Nvidia Game Driver 512.59} (3060 Ti LHR rig)
Hi Joanna,
Thanks for your help!
I actually figure it out myself last night. I just turn mclock down to 800 and don't set limit for cclock, it works fine. And then I increase mclock 50mhz step by step until it crash. Finally it seems it's stable now. [image: image.png]
Thanks, Michael
LadyJoanna @.***> 于2022年6月6日周一 05:35写道：
New issue for me today everything was fine for months, hashrates declined overnight and gpus reset to default settings.
The fix was drastically reducing mclock, I was at 1300/1100 and settled on 800mhz. My power limit stayed the same around 60% and cclock is still locked at 1380. Hashrate obviously decreased around 3-5mh but efficiency is around the same so nothing too dramatic. (These are my personal cards may not be the same for yours)
Good luck!
Notes: {NBminer 42.2} {Nvidia Game Driver 512.59} (3060 Ti LHR rig)
— Reply to this email directly, view it on GitHub https://github.com/NebuTech/NBMiner/issues/905#issuecomment-1147254464, or unsubscribe https://github.com/notifications/unsubscribe-auth/AEYK4YPIB4LYAOASGIXLFALVNXA43ANCNFSM5X57GO4A . You are receiving this because you authored the thread.Message ID: @.***>
Tried many times, still get this same issue.
I used 42.2 windows version, 512.15 driver version.
NBMiner i head the same issue, what i did is uninstall the new nvidia driver and installed the old nvidia driver 512.95 Release date Tue May 24 2022. After i download the NVIDIA PROFILE INSPECTOR [2.3.0.10] . https://github.com/Orbmu2k/nvidiaProfileInspector/releases
After download the nvidia profile inspector 2.3.0.10 open and go to COMMON 5 and change the following ( CUDA-FORCE P2 STATE to OFF) ( POWER MANAGEMENT MODE to PREFER MAXIMUM PERFORMANCE)
and click apply changes. "A REMINDER" if your pool miner have a secondary address change that as well it may be the pool miner address causing the error for my and i change it to the secondary one. I did it yesterday AUGUST 2 2022 and today is AUGUST 3 2022 and is running wall and no error yet for 24 hours.
NBMiner i head the same issue, what i did is uninstall the new nvidia driver and installed the old nvidia driver 512.95 Release date Tue May 24 2022. After i download the NVIDIA PROFILE INSPECTOR [2.3.0.10] . https://github.com/Orbmu2k/nvidiaProfileInspector/releases After download the nvidia profile inspector 2.3.0.10 open and go to COMMON 5 and change the following ( CUDA-FORCE P2 STATE to OFF) ( POWER MANAGEMENT MODE to PREFER MAXIMUM PERFORMANCE) and click apply changes. "A REMINDER" if your pool miner have a secondary address change that as well it may be the pool miner address causing the error for my and i change it to the secondary one. I did it yesterday AUGUST 2 2022 and today is AUGUST 3 2022 and is running wall and no error yet for 24 hours.
Unfortunately, this solution didn't work for me=(
