Quote from: uncommonBG on January 24, 2018, 07:56:36 AM

Quote from: mitserdt on January 23, 2018, 03:06:29 PM

Quote from: uncommonBG on January 22, 2018, 09:59:47 PM

Hello, guys.

I'm having issues with my 6 RX 580 Nitro+. The problem is the last few days I started getting "GPU hangs in OpenCL call" error. I did lower the clocks and I wasn't getting errors in HWiNFO64, but the "GPU hangs in OpenCL call" error still occured today, even without getting GPU errors.

Do you guys have any idea what else can cause the issue?

I've got that too, but it seems intermittend. Sometimes it goes for days without it and sometimes i've got it a couple of times a day. A reboot fixes it, also without any errors. It has been sugested that it can be triggerd by using a VNC server to access your Windows host, but I can't confirm that myself.

Well, after the last "GPU hangs in OpenCL call", I did the opposite this time, kept the clocks pretty much the same, but lowered the voltages a lot(to about 860) and now I'm getting 180 mh/s and the rig has been stable for 2 days now. I do get some CPU errors in HWiNFO64, but as I said, everything has been very stable. This doesn't really make sense to me, but it seems to work for now. However, after I get home I will check all the risers and maybe re-install the drivers.

Edit: I got the "GPU hangs in OpenCL call" last night. Can somebody please advise me what should I do in this situation?

Start with default memory clocks first then add 25 when stable until you crash again. Dial back down then thats your sweet spot. You should do this to all cards individually as no card has the same ASIC quality.