3080 black screening the computer

[Tal9954]

Hi everyone,

 

I have a problem with a specific card (gigabyte 3080 gaming oc 10g) that causes my screens to go black which can only be "fixed" by restarting the computer.

the problem is 100% the gpu because the specific card is causing the issue on 2 different builds. ill state my build here, as im not certain on the other person's rig

 

my rig-

cpu-AMD Ryzen 9 3900X

ram- Corsair DDR 4 16G (8Gx2) 3600 CL18 Vengeance RGB 

PSU-Corsair HX1200 1200

Mobo-X570 Aorus master

windows 10 pro 64 bit 22h2 ver.

 the black screen happens most of the times while gaming on not demanding titles ( rocket league, dota 2 etc.)

the problem started on my friend's pc, which is the original buyer of the card,  and after a little troubleshooting it seemed to be fixed, i did these steps to try and fix the problem

 

-Reinstall drivers using DDU 

-tried to recreate said problems using furmark and occt power running at the same time (couldnt recreate said problem)

-updating mobo bios to the latest version.

-reseating the gpu.

 

after all of that, the problem seemed to be solved  for a few weeks just for it to appear again.

after it appeared again, i took his gpu and replaced it with mine 2080 ti as i have more time to tinker around with the faulty gpu.

 

when i hooked the card to my computer it seemed to work with no problems so i figured the problem might have been with his gpu+psu combo.

 

as of today after 2 months, the problem started to appear on my pc aswell.

 

it seems like the problems start to happen after a while of the card being reseated on both computers.

also to note, it seems like sometimes when it happens the "Display driver nvlddmkm stopped responding and has successfully recovered." event on event viewers pops up (but not always.)

 

i thought it might be a thermal related problem but its weird that its only starts after a while of being reseated.

 

 

some things to note:

1) card should be under warranty but the buyer bought it from newegg and shipped it to Israel, which result in newegg not taking responsibility for the card and gigabyte Israel not taking responsibility because in here the warranty is dealt by the stores in which you bought the card. (so the card is not in warranty basically )

2)i have some knowledge on opening gpus before, so i dont have a problem doing that if you think that thermals might be the issue.

 

lastly i want to apologize on any formatting issues or stuff that are harder to read, as my english is not the best.

 

if i have any missing information, please let me know so ill quickly add it to the original post.

 

 

 

[PDifolco]

Unless it's some defect on the GPU it should come from either high thermals or clocks going too high vs voltage

In non demanding games the second case may be more probable, so

- Check GPU temps /clock/voltage while gaming

- if no thermal issue but very high clock try overvolting (slightly) your GPU

 

[IkeaGnome]

How old is this card? Was a support bracket used for it?

[Tal9954]

49 minutes ago, IkeaGnome said:

How old is this card? Was a support bracket used for it?

the card is about 2 years old. and no support bracket was used for it no

[IkeaGnome]

19 minutes ago, Tal9954 said:

the card is about 2 years old. and no support bracket was used for it no

Mind taking pictures of both sides of the pcie fingers and the lock where it plugs into the motherboard?

[Glazarus]

Have you checked the windows log? The problem seems similar to the one I have with my 4090 and the BSOD isn't visible since the GPU crashes, but in the Windows log you can find clues. If nvlddmkm.sys is mentioned, then it's probably the same kind of error. 

 

This seems to be solved by using DDU and re-installing Nvidia drivers. However, the issues can return, even after a while. I thought I fixed the problem, but after a month it started happening again and I'm once again going through tedious support suggestions of re-installing the drivers and whatever. 

Clues point to the GPU not able to access nvlddmkm.sys properly or that the latest drivers are bs. It might even be that Windows is acting stupid in conjunction with driver bs. 

So before moving around the hardware, check Windows logs and see if there are any hints to the crashes and what files were involved with them. 

[Tal9954]

Okay so a little update: I moved out of my house so I didn’t touched my pc until yesterday where the pc black screened in a simple game called cult of the lamb

 

ive tried to use DDU and to my surprise I’ve got black screen while ideling aswell

 

im pretty certain it’s not a temp issue

although while opening the card and putting fresh thermal paste I did discover that one of the vrams thermal pads is dry and crumbling 

 

not sure if that’s the issue and I would like to hear more since  I can’t even troubleshoot in windows anymore 

[ebu_talha]

i have the same problem:

- GPU rx 6700 xt

- CPU ryzen 5 7600x

- 650W psu

- DDR5 RAM 6000 Mhz 32 GB

- GIGABYTE motherboard b650 x ax

i tried everything ddu delete and install drivers, reinstall windows, unplug and plug back the cabels. i am tired i dont have an idea what is happening all the parts are brand new everything is new