• Happy holidays, folks! Thanks to each and every one of you for being part of the Tom's Hardware community!

[SOLVED] 1080Ti purple spots and bsod

Toggle sidebar Toggle sidebar
Steffenbk1

Steffenbk1

Distinguished
Feb 4, 2014
49
3
18,545
My 1080ti has started to crash(BSOD) on me, but it feels very random at times. Sometimes it wont crash at all when playing some games, other times it crashes every time on boot, so it feels very random.

Things ive tried:

  • New cooling past
  • Electrical clean around the board
  • Clean install of nvidia drivers and tested multiple version
  • Clean installed windows multiple times
  • Memtest
  • Underclock gpu
  • Furmark bench, sometimes it can run other times it crashes on boot(BSOD)


Then i tested another GPU a 1070 from a friend and the pc has no issues now..

What can i try to do as diagnose the problem on the board?

In the state the GPU market is in now, id rather try and save this rather than buying a new one.




https://imgur.com/p8XWLqw
View: https://imgur.com/p8XWLqw
Purple dots

https://imgur.com/fpC1qDB
View: https://imgur.com/fpC1qDB
Purple dots

https://imgur.com/XYLP0lE
View: https://imgur.com/XYLP0lE
Purple dots

https://imgur.com/Fg1Piaq
View: https://imgur.com/Fg1Piaq
Bsod faults
 
Solution
Phaaze88
I see a mixture of dust and oil, but not corrosion.
That's not the memory you've taken pics of, but the power delivery/VRM. The memory is on the opposite end of the gpu, surrounding the gpu die.
Sort by date Sort by votes
Image #3: Looks like the 'space invaders' artifacts. It's related to the card's memory. I would like to suggest you refrain from pushing the gpu with Furmark.
-Clean install gpu drivers using Display Driver Uninstaller. You may have already done that - then skip this.
-Change video ports. If you're getting video through HDMI, try DP, or vice versa.
-Is the memory overclocked? If so, remove it, or if you have an application applying the OC at Windows startup, disable it.
-If the above doesn't work, lower the memory clock by a few hundred points or so.

If none of those help, then it's highly likely that nothing more can be done.
 
  • Like
Reactions: dotas1 and Deleted member 2731765
Upvote 0 Downvote
Which PSU/power Supply unit have you been using to power up your GPU ? From those screens, it looks like a GPU VRAM corruption issue, though I'm not jumping on any conclusion right now. What is the GPU TEMP value while gaming or stress testing any app ? Did you overclock your system memory as well ?

Those artifacts do hint that the GPU might be faulty, but can you test this card on some other PC ?
 
  • Like
Reactions: dotas1
Upvote 0 Downvote
Phaaze88 said:
Image #3: Looks like the 'space invaders' artifacts. It's related to the card's memory. I would like to suggest you refrain from pushing the gpu with Furmark.
-Clean install gpu drivers using Display Driver Uninstaller. You may have already done that - then skip this.
-Change video ports. If you're getting video through HDMI, try DP, or vice versa.
-Is the memory overclocked? If so, remove it, or if you have an application applying the OC at Windows startup, disable it.
-If the above doesn't work, lower the memory clock by a few hundred points or so.

If none of those help, then it's highly likely that nothing more can be done.
Click to expand...
ah yes used DDU to uninstall, so tried that indeed.
But using DP on two monitors, could try HDMI and such, however i've never overclocked this gpu.
And ive tried to underclock the gpu as well with msi aft. Went down 100 points each crash to the lowest but to no avail..
 
Upvote 0 Downvote
Metal Messiah. said:
Which PSU/power Supply unit have you been using to power up your GPU ? From those screens, it looks like a GPU VRAM corruption issue, though I'm not jumping on any conclusion right now. What is the GPU TEMP value while gaming or stress testing any app ? Did you overclock your system memory as well ?

Those artifacts do hint that the GPU might be faulty, but can you test this card on some other PC ?
Click to expand...
PSU is a Corsair TX750M, and ive not overclocked the gpu or cpu ever on this rig, or ram but im on XMP II profile. GPU temps with furmarks was steady around 70-75 when it worked and gaming it had a peak of 80 but was around 70-78 average on heavy games.
 
Upvote 0 Downvote
That's a good/decent Power Supply unit from Corsair though. Shouldn't pose an issue as such, unless it is seriously aged. I still think your 1080 Ti GPU is faulty.
 
Upvote 0 Downvote
Steffenbk1 said:
And ive tried to underclock the gpu as well with msi aft. Went down 100 points each crash to the lowest but to no avail..
Click to expand...
That's memory clock, right?

But using DP on two monitors, could try HDMI and such, however i've never overclocked this gpu.
Click to expand...
Try a different video port, please.
If you bought an aftermarket model, almost all of them have some degree of core clock OC already applied. They don't touch memory clock though.
 
Upvote 0 Downvote
Upvote 0 Downvote
Phaaze88 said:
Wow, that's real nice! 👍
Click to expand...

They just got back to me, and said they've done furmark for 24 hours and 3dmarks tests and found no issues... I'll go to a friends house and test it when i get it back, but im currently borrowing a 1070 and i have no issues what so ever though. Could it be a power problem from my psu?
 
Upvote 0 Downvote
WHAT... that smells of bull, and yet you can only take their word for it...
I am not familiar with the psu causing the issues you've seen... maybe someone else is, but I am not. I'm just one guy though.
BSODs can be from drivers(incompatible/corrupt/missing), unstable overclocks(cpu or ram), and even overheating.
Those little artifacts: gpu memory. Unstable overclock, overheating, or it's on death's door...
 
Upvote 0 Downvote
Phaaze88 said:
WHAT... that smells of bull, and yet you can only take their word for it...
I am not familiar with the psu causing the issues you've seen... maybe someone else is, but I am not. I'm just one guy though.
BSODs can be from drivers(incompatible/corrupt/missing), unstable overclocks(cpu or ram), and even overheating.
gpu memory. Unstable overclock, overheating, or it's on death's door...
Click to expand...
These are the pics they sendt me.. However i notice in the furmark they're running in 1080p.. i didnt test that though since i did it in 4k with dport, so maybe they used hdmi
pjzthIm.jpeg
rtifacts:
9iU8mLd.jpeg
 
Upvote 0 Downvote
Steffenbk1 said:
These are the pics they sendt me.. However i notice in the furmark they're running in 1080p.. i didnt test that though since i did it in 4k with dport, so maybe they used hdmi
Click to expand...
You normally play at 4k?

The screenshots themselves aren't very useful.
The TimeSpy screenshot: "Ok, how do I know if artifacts didn't appear during the run... did someone really sit there for 24 hours and watch for artifacts? Was a recording taken, in which someone reviewed for said artifacts?"
The Furmark shot is more of the same thing. I can't interpret anything from it.


Your BSODs and artifacts are possibly 2 separate issues, and they managed to dodge both...
Backtracking a bit:
1)You've clean installed Windows, but after that, you went to the motherboard manufacturer's site to update chipset, lan, and audio drivers, etc? Used Nvidia.com for gpu drivers?
2)How old is that Corsair TXM?
3)What chassis is the hardware housed in? The folks doing the RMA likely ran the card on an open bench.
 
Upvote 0 Downvote
Phaaze88 said:
You normally play at 4k?

The screenshots themselves aren't very useful.
The TimeSpy screenshot: "Ok, how do I know if artifacts didn't appear during the run... did someone really sit there for 24 hours and watch for artifacts? Was a recording taken, in which someone reviewed for said artifacts?"
The Furmark shot is more of the same thing. I can't interpret anything from it.


Your BSODs and artifacts are possibly 2 separate issues, and they managed to dodge both...
Backtracking a bit:
1)You've clean installed Windows, but after that, you went to the motherboard manufacturer's site to update chipset, lan, and audio drivers, etc? Used Nvidia.com for gpu drivers?
2)How old is that Corsair TXM?
3)What chassis is the hardware housed in? The folks doing the RMA likely ran the card on an open bench.
Click to expand...

My desktop res is at 4k and i normally play everything on 4k with medium to high settings depending on the game. However it did crash a lot playing only heartstone though in window mode.

Yeah it look quite odd how they just took pics like these and sendt it, very unprofessional it feels like. On top they want me to pay 60$ for the diagnostics... The card is on the way back to me now, so ill ask two friends who is running currently a 3080 and another friend with a 3080ti if i can test the gpu in their riggs to see if it bricks.

  1. yes updated all drivers from the asus webpage, got the latest gpu driver. Also did clean install to test other drivers with DDU
  2. PSU is 3 years old
  3. my case is a fractal r6
 
Upvote 0 Downvote
Phaaze88 said:
Gonna have to wait and see what your buddies find, cause I don't know at this point.
Click to expand...


Took me sometime to go about this again. But metup with my friend had it tested. He dosnt get the pixel symptoms that i get, however we did get a bluecreen and some other issues like it going just black and rebooting. But we got the same BSOD faults. Tested it today back on my end after using a 1070 with no issues, and i cant even get chrome to open before it dies..

First image is my friends
Second image is my bsod faults


0xFzp33.png

eli8oMe.png
 
Upvote 0 Downvote
You must log in or register to reply here.

TRENDING THREADS

Latest posts

Moderators online

Share this page

COMPANY
RESOURCES
FOLLOW
Top Bottom