Hello
How i identify wicht gpu is failing?
My rig is 6x r9 390x
Plug hdmi cable, run stress test with stock clocks, if you see artifacts , sit, try not cry,roll,cry a lot
Hahaha another way?
The gpu mining for a few hours, them fail open cl call and stop mining , them after a time, gpu freeze w10 and not signal.
Claymore show in log files that gpu 4 is the problem, vut i dont know what is gpu 4
I have a same problem ... But only with Claymores Dual miner . I checked all card separately, only one card was connected to the mobo . all card worked fine as a single card . As soon as I plugged in all card and started the miner , first I see gpu 4 is hashing around 4 mh/s few minutes later it was hashing with 0 , under wattman settings , everything was grayed out , was not able to change any settings ... Quick restart settings become available , as soon started the miner its grayed out again .
Used another miner program , that was the only way for me to fix the problem .
My setup ( 6xrx480 ) was hashing fine for a few hours with another miner , then I switched back to claymore miner and everything was ok again .
this problem never happened when I used the exactly the same cards in one setup . One of my rig have 6 cards from differently manufacturers with different default clock ... On that setup I see problems with gpu 4 almost every day .
Maybe is a bug ?
I had similar issues with 2 PowerColor Red Devil 470s. In Win10 they get bugged sometimes after restart, only way to fix is to either
Disable Crossfire
Enable, then Disable crossfire.
Before I found this, I would reinstall driver, after DDU, which usually didn't work, and was tedious. You can try changing slots so it doesn't try to crossfire them? Annoying issue I only have on Win10. The clocks get stuck at 300/300 which gets ~4MH.
I don't have the issue on Win7, though some cards don't let me change some Wattman settings while miner is active. Occasionally I need to restart.