Try raise VDDCI / MVDD for card with Bus ID 0b:00 (GPU5 according overview tab). It’s seems too low
Possible this card can’t stay stable at 700 mV.
Try raise by step of 12mV like 712, 725, 737, 750 and so on …
OK. Thank you.
btw, why is 12mV?
to be more correct
1 step is 6.25 mV but driver operates only with integer values (6 mV) and for my experience only 2 steps changes matter. S
fixed!
it’s riser card issue.
thank you
bad riser can bring real pain 
Anyway glad to hear that you found & fixed it
I have the same problem. how you solved it
I think this issue can be attributed by high OCs too. In my case I had 5700 XTs on NiceHash, each time the Dag is loaded (i.e. Loading another algo for profitability) this happens; dropping the OC solved the issue.
Something I noticed was that if you have a relatively high rejected shares it means your OCs are unstable and you might run into some issue; in my case this particular issue.
Hope this helps in any way.
I have the same issue, but my 6 x Radeon 5700 XTs work fine during 3 or 4 minutes and the I get the error on 2 of them:
Message issuing precheck failed and SMU might not be in the right state.
The OC is pretty low in Hashrate terms
Can anyone help?
Same issue, very conservative OC settings.
On 0.6 191@210109
Would greatly appreciate any insight
same issue for me, any updates with that please?
I am facing the same issue, I changed my riser to one VER09, hope this helps because I have made a lot of adjustments and nothing works. Thanks.
Update: None issue presented so far, apparently it was my riser because my my rig did not work more than 1.5 days without the reboot issue “teamredminer reboot”.
I hope this helps- I just went through hell and back with this issue. Mobo is B250, I had 7 5700s running on it and I wanted to bring over 2 6900s since the latest firmware added support for big Navi cards… Yeah I never got the 6900s up and running with the others. On the last of a dozen attempts, the PSU literally went up in smoke. I’ve never seen so much magic smoke. I pulled the 6900s off, machine would boot but ALL the cards would throw the smu may be in wrong state… Message. After 6 hours of trial and error, swapping mobos, cards, risers etc… what ended up working on the end was two things:. REMOVE ALL OC, RESET BIOS by removing battery. I was suicidal because I thought I burned 8k in GPUs… Back to good now, minus a PSU.
2 Likes
When you say reset bios by removing battery, I just want to confirm that means reset motherboard bios.
Yes. Pull the Mobo battery for a minute or two.
I tried this and it didn’t help. I tried replacing risers and still had the issue. Just out of curiosity, I pulled the card in question and the problem persisted. I started pulling other cards one by one and the issue persisted. It was always the first GPU that was in my PCI-E tree that was throwing the error. I had this problem on two of my rigs. What I ended up doing was saving my OC settings as an OC Template, removing the miner from my farm, wiped my SSD and started it back up with a clean USB Flash Drive of hive with farm hash and same rig name. Once it started on first boot, issue was gone. Then I just applied the OC Template to the rig and was backup up and running in 5min. Worked on both of them.
My guess is the Powerplay table that is saved got corrupt. Next time it happens I’ll try to delete just the PPT file and see if that fixes the issue.
Hope this saves others some time from banging their head up against a wall like I was.