F5F Stay Refreshed Hardware Desktop Graphics card crashes

Graphics card crashes

Graphics card crashes

H
HolgerHolm95
Junior Member
21
10-19-2022, 02:48 PM
#1
Hello, Your system details: Z790 motherboard (ASRock ASRock Z790 Lightning WiFi), Intel 12900k, Zotac RTX 3080 Trinity OC CORSAIR DOMINATOR PLATINUM RGB DDR5 RAM 32GB (2x16GB) at 5600MHz, C36-36-36-76 1.25V Powerspec 1250W PSU, Samsung 970 EVO, Windows 11 Pro 64bit. Issue: After some time (5 to 45 minutes), the display starts tearing before a crash appears with a black screen. On restart, the motherboard fails to recognize the graphics card until you remove it, then boot using onboard graphics and later reinstall. Crashes occur during gaming, benchmarks, stress tests, and even while browsing. You've tried: swapping motherboard, PSU, reinstalling OS, underclocking, adjusting GPU power limits, reapplying thermal paste, using a different PCIe slot, changing BIOS/drivers, disabling XMP. Event Viewer logs show frequent NVDLM errors with various IDs. Updated July 26, 2024 by Gothicwolf Card crashed after some time in multiple PCs.
H
HolgerHolm95
10-19-2022, 02:48 PM #1

Hello, Your system details: Z790 motherboard (ASRock ASRock Z790 Lightning WiFi), Intel 12900k, Zotac RTX 3080 Trinity OC CORSAIR DOMINATOR PLATINUM RGB DDR5 RAM 32GB (2x16GB) at 5600MHz, C36-36-36-76 1.25V Powerspec 1250W PSU, Samsung 970 EVO, Windows 11 Pro 64bit. Issue: After some time (5 to 45 minutes), the display starts tearing before a crash appears with a black screen. On restart, the motherboard fails to recognize the graphics card until you remove it, then boot using onboard graphics and later reinstall. Crashes occur during gaming, benchmarks, stress tests, and even while browsing. You've tried: swapping motherboard, PSU, reinstalling OS, underclocking, adjusting GPU power limits, reapplying thermal paste, using a different PCIe slot, changing BIOS/drivers, disabling XMP. Event Viewer logs show frequent NVDLM errors with various IDs. Updated July 26, 2024 by Gothicwolf Card crashed after some time in multiple PCs.

J
Johnny47751
Junior Member
43
10-19-2022, 04:26 PM
#2
You might attempt that setup on a different machine, or ask a friend for their PC if you don’t have one. If it doesn’t work there too, it’s likely faulty. To be cautious, try using another GPU in your current system to check the outcome. Do you still have coverage for the 3080?
J
Johnny47751
10-19-2022, 04:26 PM #2

You might attempt that setup on a different machine, or ask a friend for their PC if you don’t have one. If it doesn’t work there too, it’s likely faulty. To be cautious, try using another GPU in your current system to check the outcome. Do you still have coverage for the 3080?

L
LOLWHYUSOSAD
Member
52
10-20-2022, 05:13 PM
#3
It failed after two months of warranty. I tested it yesterday on another PC with heavy usage—Doom at maximum graphics settings, Furmark knot glitches, etc.—it ran for a few hours but crashed eventually. I suspect a hardware issue; first thought it was a memory chip, but Doom worked fine without it, using all 10GB without problems. There might be thermal problems, with temperatures reaching high 80s and the board staying low 80s, plus a power draw of around 320W.
L
LOLWHYUSOSAD
10-20-2022, 05:13 PM #3

It failed after two months of warranty. I tested it yesterday on another PC with heavy usage—Doom at maximum graphics settings, Furmark knot glitches, etc.—it ran for a few hours but crashed eventually. I suspect a hardware issue; first thought it was a memory chip, but Doom worked fine without it, using all 10GB without problems. There might be thermal problems, with temperatures reaching high 80s and the board staying low 80s, plus a power draw of around 320W.

H
Hyper_B0Y
Member
215
10-22-2022, 10:38 PM
#4
It should begin to slow down when temperatures rise too much. It seems similar to just above 90 degrees on the 3080 model. I believe the issue isn't primarily temperature, because it also fails during light use. Disappointing news—it’s now outside the warranty period.
H
Hyper_B0Y
10-22-2022, 10:38 PM #4

It should begin to slow down when temperatures rise too much. It seems similar to just above 90 degrees on the 3080 model. I believe the issue isn't primarily temperature, because it also fails during light use. Disappointing news—it’s now outside the warranty period.

D
dawgal
Member
126
10-23-2022, 03:27 AM
#5
I doubt thermals are the main issue since many users mention thermal crashes and problems with Zotac cards, especially in cold conditions. Adding a waterblock might help temporarily, but I’m hesitant to spend $200 just to see if it resolves the problem.
D
dawgal
10-23-2022, 03:27 AM #5

I doubt thermals are the main issue since many users mention thermal crashes and problems with Zotac cards, especially in cold conditions. Adding a waterblock might help temporarily, but I’m hesitant to spend $200 just to see if it resolves the problem.