3080 – GPU Driver Error, no temps

[JDD]

Guys, I have been having issues with my Gigabyte 3080 Gaming OC.

What I have done…

• 1. I have moved the cards to a separate RIG configuration (for more airflow). There are only two cards on the RIG for now
• 2. I have changed the Thermal Pads on both cards – I have no issues with Thermal Throttling
• 3. I have upgraded to the newest Nvidia drivers – 460.67. Before moving them to a separate machine they had been working on 455.45.01 – Same issue
• 4. I have lowered the Overclock settings – As you can see I have brought the MEM and Core to a minimum
• 5. I have read a couple of posts that suggest to take away the Overclock – the issue that arises is that they start consuming 320 – 300 Watts, which is a lot.
• 6. I have stopped AUTOFAN and manually turned on the fans to a constant 90 % for better cooling

All of this has brought me nowhere… Can someone with those cards share his experience if he has the same issue?

My issue is that both cards are working below 50 Degrees constantly and they still have issues with the temperatures. I have read that the VRAM temperature (Junction temp) of those cards is very high running at an average of 100C, nevertheless I don’t see how that is the issue, as HIVEOS cannot detect those specific temps of the cards….

[Hrvoje_Soba]

Hi, did you change both backplate and top thermal pads or just backplate? I have added backplate thermal pads and lowered the temps a bit, but it still throttles, so I am about to change the factory top ones. I have no other idea what to do with it.

H

[Hrvoje_Soba]

Here is my nightmare with gigabyte oc cards. Evga works great from day one, Gigabyte works great 1-2 days, then restarts 20 times a day etc, dissregarding room temperature. I have added backplate thermal pads, but it`s still not there yet. Any idea for setting these cards? Maybe another bios?

[JDD]

hey,
I have changed the thermal pads on the backplate and all of them on the memories, I still have problems with them… What cards are you using? If you have Gigabyte Gaming OC you will need pads 1mm and 2mm for around the core. I have put pads on all of the VRAMS you can watch a video from a guy called Son of Tech, he does it on his Aorus master 3080 but the cards are identical on the inside… for me it didnt help it still stops working i just stabilized the MH/s

[JDD]

From the picture that you have sent I can see that all of your Gigabyte cards are Thermal Throttling for sure. Easiest way to see that is when they cannot reach the maximum Power Limit that you have given on the Overclock…

[JDD]

You have given a PL to 250 W - you see that they vary between 203 and 219… THE EVGA cards as you can see Power Limit 250 the card manages 250, so its fine…

[Hrvoje_Soba]

I am waiting for 2mm thermal pads to arrive to change it, but funny thing is that sometimes they work great for a few days, sometimes I can`t get them to stabilize for days.

Which kind of settings should I go with?

[Tincho]

Hi, Did you try downgrading the HiveOS version?, I think I solve it like this

[JDD]

Hey Tincho,

Nope I havent tried with other versions. Could you please tell me, which version is stable for you, so I can downgrade…

[Tincho]

it did not work again after a day it failed again

[JDD]

same for me… i still have the issue… I read somewhere in the forum that if you put the RIG on Windows the issue doesnt exist…

[Hrvoje_Soba]

What about gpu bios change? Anyone tried that?

[Hrvoje_Soba]

I have menaged to have them not restarting, but had to lower the oc to 1000 mhz which gives around 91 mh rock stable. Due to this, Im sure its regarding overheating. Would installing a better cooler help?

[Elscreauwto]

no idea if it helps; but I had the same error with my 5700 xt’s and for me it was related to the way I powered the risers. I used SATA to 6pin, but changed it to directly 6pin from the PSU what solved the problem (never seen this error since)

[fishcake]

same problem here. are hiveos devs even looking into this? there surely has to be some reason as to why it’s happening.

[jmw60546]

I am having a similar issue with one of my 3080 cards. I did an inspection and noticed to +2 pin on one of the 6+2 PSU connectors was loose. I reseated it and testing now to see if it makes a difference. Typically the rig would run fine for 2 hours then reboot or go unresponsive. The only error showing is the GPU Driver Error, no temps.

[UncleSamThe1st]

Any updates? I have the same problem, tried rolling back Drivers, decreased some OC settings, changed the miner, still nothing changed, every 4 hours or so i get the no temp error and reboots.

[JDD]

Hey guys,

so I am still having the same issues, nothing has worked for me as well. The only option is to decrease my OC even more, but consequently the hash rate drops a lot, which isnt an option for me… I have no idea what more to try

[jmw60546]

Mine has been stable for the last 18 hrs. After re-seating the +2 GPU connector, I dropped all the OC settings down and increasing them slowly every 4 hrs without issue so far. I have been running Core -200, Mem 800, PL 235 for the last 9 hrs without issue. I just bumped Mem up to 900 and will see how it goes. Getting 92Mh/s right now. Hoping to get it to 95-96 and stable but will take what I can get.

[fishcake]

GPU driver error, no temps (Auto fan disabled)

this is such a joke. opened an issue on github but hiveOS devs don’t even care [Screenshot 2021-04-09 at 13.27.08] hashrate watchdog (and autofan) is also broken [Screenshot 2021-04-09 at 13.28.12] I am about to switch platforms.