GPU driver error, no temps (Auto fan disabled)

[killopatra]

I’m on 0.6-203@210519. My driver rollback hasn’t been perfect, my rig did reset yesterday. I can’t be certain what the issue is. My rig reset during the hottest part of the day but my core temps never get above low 50°s so I’m not sure if heat is the culprit. I’ve lowered oc’s considerably and still gotten this error so I’m not sure my oc is the issue either. I will say my rig has been more stable when mining on octopus algo vs ethash so maybe ethash is having some affect? I’m really not sure.

[Devilz]

Mine reset also, around 16 hours miner running. Checked logs, miner don’t show any error. I’m thinking reflashing my os to earlier version.

[mineurdefonds]

Same problem for me… really think about a hiveos issue…

[ReScUe]

Tested now few things:

Test Rig: 2 Cards
→ Changed Only mem clock. Works for now.

1060 Rig with 6 Cards:
→ Changed mem settings to same Value (1000)
Cant set fan speed → Phoenixminer reboot

The weirdest thing about it?
The error does not come because an overclocked card. Its coming because of an Stock Card with only Powerlimit (wich worked 3 Weeks long without an issue).
After changing the Settings of the other Cards its not working anymore → Now after an reboot an other card has the failure now its one of the oc cards.

I dont know whats wrong tbh. Running in circles.

Greatings.

[mad]

how to know wich driver is the stable one for my rig i have 2x3080 and 4x3070

[MinerMH]

@ReScUe

used gpu find - located faulty GPU

Hi ReScUe!
What di you mean exactly with “ppu-find”?
Can you explain it a bit more how this works?

Have HiveOS current version.

And also having this troubles with GPU Driver error, frequently between 16h and 2h uptime.
Got any solution in this problem?

THX KR MMH

[ReScUe]

Hei,

you can use command gpu-fans-find to find the faulty GPU with his ID.
If its GPU 1 you could use gpu-fans-find 1 to find the specific GPU in your RIG.

Fans of this GPU will spinn up - thats how you find it.
Then you could check your riser → Use an other one. Maybe its working.

I didnt find a solution, only a workaround wich is working for some of my GPUS:
If im only setting mem clock and not core its working. (But only for some…)

One of my Rigs worked fine, after reboot same failure. Could not find something so its running without OCs…

Hope i could help you a bit.

Greatings
ReScUe

[MinerMH]

@Greenhadouken

Unfortunately, it only ran stable for like a week. Now it’s back to giving me the same error. Sorry. I was really hoping I solved that issue for the both of us.

Hi Greenhadouken!

I’m having this failure sometimes 1/hour, sometimes 1/12 hours.
Did you ever find a solution for this?

Thanks a lot!
KR MMZ

[MinerMH]

@cmrho

I fixed this error (GPU driver erro, no temps) by performing the following on my Nvidia rig:

• enabling maintenance mode (don’t load drivers) in the Settings tab.
• rebooting
• running ‘nvidia-driver-update --list’ command and selecting a previous stable driver version. Let it install. I’m guessing AMD cards have the same ability.
• clearing all OC settings in the Overclocking tab
• reboot, confirm mining success with no OC settings.
• once confirmed, setup OC again as you would a new rig with measured changes.

Hi cmrho!

Pls could you or somewho else explain in simple steps how to do, especially point 1 and 3.
Pls do this in an easy way for newbies! Thank you very much!

I’m using HiveOS in current version and Biostar BTC-360 Pro.
1x 3070, 1x 3080, 3x 5700 XT

THX KR MMH

[ReScUe]

Hi MinerMH, there are good guides from HiveOS on their Knowledgebase:

Point 1: https://hiveos.farm/guides-driver_upd/ at the end its explained.
Point 2: Click on the Power Button then select Reboot
Point 3: Use Hive-Shell: https://hiveos.farm/guides-hshell/ then https://hiveos.farm/guides-driver_upd/

Link to Knowledgebase: https://hiveos.farm/knowledge-base/

Greatings
ReScUe

[MinerMH]

THX for your answer!

In the meantime I could fix this issue by lowering the OC at the RTX 3080.
Maybe OC too high this is one of the most made failure! (maybe)

Also AutoFan works properly now since then…

KR MMH

[ReScUe]

Hei,
Jeah thats the other Point of Failure.
For me it didnt fix the Problem…

But nice that it works for you

Have a great Day!

Greatings
ReScUe

[Jamespigeon]

This is still an issue, its too bad that Hive dont fix this…

[hfcartel]

Well in my case issue was not in hive os at all.
i figured out that my USB connected to riser is actually not plugged well. I changed USB cable, and no more this issue for me.

[Kyawhtetaung]

How to fix it

[hfcartel]

In my case i found out what caused this problem - power supply!
First I tought it is riser USB cable, but after 12 hours i got same error.
I have corsair 1200w, 1050 TI, 3070, 3060TI and 2x 3080. When i touched it - it was so hot.
It is too much GPUs for 1200 W power supply.
I added one more power, switch 3080 and 3070 to it and now everything works perfect for 3 days.

[hunters42]

hocam bende 1 adet suprim x 3080 var bende aynı hatayı alıyorum çözüm bulamadım 99.5 ile phonexminerde kazım yapıyorum daha önce trex de kazım yaparken gene aynı hatayı alıyordum bir türlü çözüm bulamadım

[hunters42]

bendede aynı sorun var msi suprim x 3080 99.5 mh ile kazım yapıyorum aynı hatayı alıyorum sistem otomatik yeniden başlatıyor ve havuzdan düşüyor nerdeyse 2 güne bir oluyor

[KryptoMc]

I had the same problem.(ERGO, 2miners, t-rex / rig:3060ti,3070ti, 3080)
I solved it by:

1. Downgrade hiveos.
2. Downgrade nvidia drivers to stable version.
3. Many tries to set good OC.

[rardniyom]

After I changed the risers, thess problems (not showing temp / fan / power intake) has gone. Maybe, I’ll consider changing the power input choices (6-pin to SATA at the bottom / at the side / MOLEX) for those risers later. Moreover, make sure you plug CPU cable modified with 2-tail 8-pin PCIe to power the graphic card into the CPU port on the power supply unit, not the peripheral ports.