Zotac 3080 Trinty OC 10GB Crashes after driver install

[vladimirc]

I've recently purchased Zotac 3080 Trinty OC 10GB LHR.
Whenever I try to update drivers past 472.12 card will crash imminently and is only accessible through safe mode or virtual adapter (parsec).
I've tried everydriver from the list, but no dice.

Card works fine on 472.12, runs every game I tried at ultra 1080p (rdr2, cyberpunk, uncharted) at high frame rates.

Card is connected to 1kw PSU (superflower ~10years old) which worked fine for ages and still (to my best knowledge) does and is plugged in Steel Legend 660 (no raisers).
PSU never served more than ~700W.

Tried flashing the card with several vbios versions but no luck. Does anyone have some idea how should I proceed?

 

[eidairaman1]

Get hwinfo64 and get a screenshot of power supply voltages going to the motherboard and gpu please

And please do not bios flash anything, it doesnt help.

I've recently purchased Zotac 3080 Trinty OC 10GB LHR.
Whenever I try to update drivers past 472.12 card will crash imminently and is only accessible through safe mode or virtual adapter (parsec).
I've tried everydriver from the list, but no dice.

Card works fine on 472.12, runs every game I tried at ultra 1080p (rdr2, cyberpunk, uncharted) at high frame rates.

Card is connected to 1kw PSU (superflower ~10years old) which worked fine for ages and still (to my best knowledge) does and is plugged in Steel Legend 660 (no raisers).
PSU never served more than ~700W.

Tried flashing the card with several vbios versions but no luck. Does anyone have some idea how should I proceed?

 

[vladimirc]

Get hwinfo64 and get a screenshot of power supply voltages going to the motherboard and gpu please

And please do not bios flash anything, it doesnt help.

Full report: https://cvele.github.io/temp/report.HTM

Computer: ASRock B660 Steel Legend
CPU: Intel Core i7-12700 (Alder Lake-S 841, C0)
2100 MHz (21.00x100.0) @ 598 MHz (6.00x99.8)
Motherboard: ASRock B660 Steel Legend
BIOS: 8.03, 11/22/2022
Chipset: Intel B660 (Alder Lake-S PCH)
Memory: 32768 MBytes @ 1197 MHz, 17-17-17-39
- 16384 MB PC25600 DDR4 SDRAM - Kingston KHX3200C16D4/16GX
- 16384 MB PC25600 DDR4 SDRAM - Kingston KHX3200C16D4/16GX
Graphics: Zotac RTX 3080 LHR
NVIDIA GeForce RTX 3080, 10240 MB GDDR6X SDRAM
Drive: WDC WD101EFBX-68B0AN0, 9766.4 GB, Serial ATA 6Gb/s @ 6Gb/s
Drive: WDC WD101EFBX-68B0AN0, 9766.4 GB, Serial ATA 6Gb/s @ 6Gb/s
Drive: WDC WD101EFBX-68B0AN0, 9766.4 GB, Serial ATA 6Gb/s @ 6Gb/s
Drive: WDC WD101EFBX-68B0AN0, 9766.4 GB, Serial ATA 6Gb/s @ 6Gb/s
Drive: WDC WD101EFBX-68B0AN0, 9766.4 GB, Serial ATA 6Gb/s @ 6Gb/s
Drive: WDC WD101EFBX-68B0AN0, 9766.4 GB, Serial ATA 6Gb/s @ 6Gb/s
Drive: Samsung SSD 980 1TB, 976.8 GB, NVMe
Sound: Intel Alder Lake-S PCH - cAVS (Audio, Voice, Speech)
Sound: NVIDIA GA102 - High Definition Audio Controller
Network: RealTek Semiconductor RTL8125 Gaming 2.5GbE Family Ethernet Controller
Network: Realtek Gaming 2.5GbE Family Controller
Network: Xbox Wireless Adapter for Windows
OS: Microsoft Windows 11 Professional (x64) Build 22621.900

 

[eidairaman1]

Change to sensors only and get a screenshot of power supply voltages going to the motherboard and graphics card

 

[vladimirc]

Change to sensors only and get a screenshot of power supply voltages going to the motherboard and graphics card

Here it is

 

[Colddecked]

Just don't update for now? could be some issue with Zotac bios and the new driver... report it to nvidia and zotac.

 

[vladimirc]

Just don't update for now? could be some issue with Zotac bios and the new driver... report it to nvidia and zotac.

It did cross my mind and I did my best to search the google for any other reports.
This did not result in anything tangible. Given how this card is pretty old now, and so is the driver version, I would assume there to be at least a question or two.

Reporting it would be logical next step.

 

[GerKNG]

did you try to DDU the drivers properly?

Download DDU (DisplayDriverUninstaller) and the latest driver.
pull your lan cable and hold shift while pressing restart.
select advanced options, starting options- restart.
press 4 (safe mode without network.)
Open DDU, wipe the drivers and restart your PC.
install the latest driver and shut the PC down.
connect your lan cable and boot your PC.

 

[vladimirc]

did you try to DDU the drivers properly?

Download DDU (DisplayDriverUninstaller) and the latest driver.
pull your lan cable and hold shift while pressing restart.
select advanced options, starting options- restart.
press 4 (safe mode without network.)
Open DDU, wipe the drivers and restart your PC.
install the latest driver and shut the PC down.
connect your lan cable and boot your PC.

Yes, did the DDU. This is outline of what I did:

- Installed the card
- Disabled onboard
- Windows installed drivers from update (472.12)
- Downloaded latest (1st of December), installed, GPU crashes with 141 code (from reliability logs)
- Changed different ports on the card: multiple DisplayPort and Hdmi
- Cleaned PCI slot with air can
- Boot into safe mode, DDU removed all traces
- Downloaded previous version of Nvidia drivers
- Repeat above loop for all drivers until I downloaded 472.12 from Nvidia website, installed, all good

On top of that flashed bios with one from archive on this website.

 

[GerKNG]

did you bought it used? (because you bought it "recently")
otherwise i would contact zotac (they are actually very nice at least in europe. i got a full refund on a 2080 ti i bought on ebay directly from zotac)

 

[vladimirc]

It was new (sealed), however it was bought in another country so I'd have to travel for warranty.

 

[eidairaman1]

Report the issue to nvidia, zotac and game makers

 

[vladimirc]

Report the issue to nvidia, zotac and game makers

Thanks. Were you able to make any kind of conclusions based on voltage data? Does it look normal ?

 

[eidairaman1]

Thanks. Were you able to make any kind of conclusions based on voltage data? Does it look normal ?

Everything is in spec

 

[es_pablo]

Did you try a clean windows installation?
I had a similar issue on AMD gpu. Ddu didn't help, formatting the computer worked well for me...

 

[vladimirc]

Did you try a clean windows installation?
I had a similar issue on AMD gpu. Ddu didn't help, formatting the computer worked well for me...

I just tried with a fresh Windows 11 installation. Temporally hooking up new sdd with fresh and de-attaching the old one. Same results.

Everything is in spec

I assume that means that PSU is working as it should ?

 

[es_pablo]

I assume that means that PSU is working as it should ?

he can't tell... i presume you're screen were taken in idle...
to evaluate if a psu is working correctly it must be examinated under heavy load.
a 10 years old psu is old, and it should be replaced, hopefully before it starts showing problems....
anyway, the best way to test and find problems is cross testing...
so, if you can, test youre gpu on another pc, and test another gpu on youre pc...

 

[eidairaman1]

Did you try a clean windows installation?
I had a similar issue on AMD gpu. Ddu didn't help, formatting the computer worked well for me...

Seems os gets easily corrupted

he can't tell... i presume you're screen were taken in idle...
to evaluate if a psu is working correctly it must be examinated under heavy load.
a 10 years old psu is old, and it should be replaced, hopefully before it starts showing problems....
anyway, the best way to test and find problems is cross testing...
so, if you can, test youre gpu on another pc, and test another gpu on youre pc...

The Input voltages to the motherboard are within atx spec, so no issue there

And just because the psu is 10 yo or 5 or 30 doesnt mean squat. The volts are there.

Also he can load test it under a game with hwinfo64 open to see what the max load is on it.

So dont spread misinformation pablo

 

[es_pablo]

Seems os gets easily corrupted


1 -The Input voltages to the motherboard are within atx spec, so no issue there

2 - And just because the psu is 10 yo or 5 or 30 doesnt mean squat. The volts are there.

3 -Also he can load test it under a game with hwinfo64 open to see what the max load is on it.

4 - So dont spread misinformation pablo

1 - volatages are not the only things a psu produce, you now? there is ripple (spikes in low and high frequences) that you can't measure with mobo sensors and pc softwares... and, i never saw a psu, even the worst, going out of specs in idle..., real tests are done under load, problems happens under load, pc freezes or reboot under load, not in idle (obiuvsly, i mean, without messing with overclock or else)
2 - so, you mean, if i use a psu correctly, and not heavy loading it, i can use it forever? tell me, what is the longest warranty an oem offer on a psu?
3 - yes, he can, he can hack the nasa center too, by doing this and that, but he seems not be awarned on what we are talking about, so, maybe, he didn't do it if we dont ask him to do and explain it to him. And i dont ask it to him becaus in my opinion reading some voltages from the mobo is not enough to test a psu.
4 - i'm doing the opposite.
here's some psu reviews with some dedicated hardware, sorry, it's not in english, but, if you know the subject, you will uderstand anyway what i'm talking about...
Thanks.