Bitcoin Forum

Hello,

in my 4 Rigs I am using 8 x NVIDIA 1060 (from different manufacturers) at each rig and overclocked every GPU a little bit using MSI Afterburner (Power Limit 65%, Temp Limit 75, Core Clock +80, Memory +750). I am getting around 24 MH/s at each card when mining ETH.

The problem is, that after some hours or days only GPUs 2, 3 and 7 (GPU range is 0 - 7) are slowing down to around 10 MH/s or even less. First I thought, that I overclocked exactly that cards too much! But then I recognized, that on all 4 Rigs it is the same behavior: Only GPUs 2, 3 and 7 are slowing down from time to time. I thought I could solve this by disabling or lowering OC settings for exactly that cards. But that does not help. The problem appears again after some time.
So I think, that any other card(s) is/are overclocked too much and slows down exactly GPU 2, 3 and 7. But why exactly always that cards? And how can I find out which cards are the problem to lower their OC setting?

Does anyone have a solution for this or maybe the same behavior? Why is this always 2, 3 and 7??

if you tried stock mem clock settings and it was not worked, increasing power limit should be work.

At stock clock settings it works without problems.
But how should I find out which cards could I overclock how much, when always GPU 2,3 nd 7 get slow and not the one which I did overclock too much!?

well, start card by card and find the sweet spots for each card.

Do you think that higher power limits will allow higher mem clock settings?

i had similar problem with my nvidia p106-6g gpus on ethos. i have 2 similar rig same gpu mobo only difference memory amount. one of 4g other one 8g. one rig crashes with same config.

after that i applied a little different oc settings on crashed rig. apprx. %3 lower but before that you must be sure psu is enough.
what about memory? i guess 4gb is not enough for 8 gpu you should insert at least 8 or 16GB

You try to increment Power Limit , in more case resolving this problem , only with board have problem .

I experienced similar issue in the last couple of months. Hope you guys can find a solution.

Increasing power limit, did not resolve the problem. It is appearing again.
RAM can not be the problem, cause other Rigs with 8 GPUs and also only 4GB RAM are running without problems.