Hey Guys if you have Claymore telling you this: Gpu1 hangs in OpenCl call,exit. this is what i did to fix it.
My Rig:
7x RX 580 8GB DUAL OC Edition on Windows 10 with Claymore 9.7
Bios Modded (only the Straps from 1750 to 2000), Crimson 17.3 Betadriver for Blockchain. 28,2MH/s /card
- I woke up to a crashed Miner.
- One of the GPU was giving out 0,000Mh/s, the rest was fine.
- Windows was giving bluescreens with (Thread stuck in device driver)
well..
what i did:
to figure out which Card was having a hickup i unpluged one after the other until i found it.
took the card out of the rig, pluged a operating one in that slot.
2.1 card worked fine in the slot that was having the problematic card in there b4 so: Its not the riser or anything behind it.put the faulty card in a slot that was operating fine(just to make sure) result: still 0,000mh/s
unpluged all the cards that did work. left the malfunctioning one in.
flashed original bios back on to the card. - didnt help
ddu - driver uninstall. back to box driver. - didnt help
went and tried several settings for overclock/undervolt etc. didnt help
had a hissy fit. - didnt help
thought the card is EX at this point. pluged the hdmi cable in, works fine ( o.0)
run a heavy benchmarktest on the GPU, worked fine.
ran the Claymore again. 0,000mh/s (repeat step 8)
let the other 6 Cards operate for 2 days and ignore the problem.
took the card and installed it on a different machine that never had a rx 580 in it (fresh driver)
0,000 MH/s for a minute or two with the old watchdog barking about the card hanging in openCL
after a minute or two: holy shit its auctually hashing at box-speed 24,4 MH/S
put the card back in the mining rig: fixed.
long story short: if DDU, new BIOS, OC settings dont fix it. Reinstall Windows or try another PC - it will fix it.
hope this helps.