This is the problem I keep having:
After reboot, display GPU or first GPU / GPU 0 will always crash after miner is started immediately as pool sends 1st job. Other GPU's are fine, but only the one connected to display, or GPU 0 has this issue and its reproducible on 2 identical rigs so I know its not a hardware issue. Then after I close the miner and start it second time, everything is fine.
Steps to reproduce:
- Reboot PC
- Start miner, wait till received job from pool, observe crash, hit H and GPU overclock is reset to default clock
- Close miner, restart miner, works fine after that
Using RX 470s, GRAFT algo.
Anyone else having similar issue? Dok is this something that could be fixed ?
Hi Dok - any idea why this would be happening?
Also I have 2 rigs with RX 470s, and on 1.8 GRAFT algo they're not stable for more than a day, crash and reboot on both. I reduced OC and increased voltage but still not stable for more than a day. Previously both rigs ran SRBMiner v1.6.8 and very stable for 30-40 days non stop - any ideas what would cause this instability on v1.8?
Thanks in advance!
About the gpu that is connected to the display, try to set a lower intensity on that one, it is using more memory than the other cards, and it probably fails because of that the first time.
It's hard to tell what could cause video driver crashes, what i recommend always is to do driver cleaning with DDU, and use 18.6.1 drivers because they are proven to be good and stable.
Then start with a lower intensity and build up slowly.