Question GPU failing under load?

Sep 27, 2023
2
0
10
Hi all,

New to the forum.

I can't seem to figure out why my pc keeps failing under load.

I have this pc for about 6 years now. Never had any issues before.

A couple of weeks ago while playing RDR2 (worked fine before) it suddenly crashed, blackscreen, still hearing some stuttering sounds from the game and a windows disconnect sound (e.g. removing USB). I reset my pc but can't. It wants to boot but gets stuck with mobo code 43 after 15 secs of 45 and 15 secs of 44, most of the time, sometimes a different code. If I hold the power button it'll turn off and when I turn it on it's fine again.

Happens after a few minutes of gameplay, sometimes later than other times. So I figured to try 3Dmark timespy and firestrike. Crashes on both at random intervals. HWinfo showed peak temp of 101 Celsius, figured thats pretty hot. So I disassembled my GPU(waterblock) checked for any damage or abnormalities and put new thermalpaste in place. Temp now max 70celsius peak temp, overal about 52 celsius under load.

That didn't change anything. Started really listening to my pc, hear some crackling noise in PSU (RM850i). Figured that might be the problem, shorting out with powerdraw or something. I return the PSU to Corsair, wait for a few weeks. Indeed PSU is faulty. Get a new one under warrenty (RM850x). Install it today, still same problem.

What I've tried so far:

Windows update
Nvidia driver update (also complete clean install removing drivers etc.)
Reset BIOS
BIOS update
Disassembling and reassembling GPU (waterblock)
Reseating my RAM
Trying different PCI slots
New PSU
15% underclock GPU

Specs

OS: Windows 10 64bit
Motherboard: ASrock x470 Taichi Ultimate
Processor: AMD ryzen 7 2700x 3,7ghz, EK waterblock with north and southbridge cover
GPU: Nvidia RTX2080 MSI Sea Hawk waterblock
RAM: Corsair CMR 16gb DDR4 3200mhz
Storage: ADATA SX6000PNP, Samsung SSD 850EVO 250GB, Samsung SSD 870 QVO 1TB
PSU: Corsair RM850X 850watt

I'm out of ideas beside buying a new GPU, but I have a custom watercooling loop so that means a lot of adjustments. Any ideas?

Greetings,
Dutch360
 
Last edited:

Zerk2012

Titan
Ambassador
Hi all,

New to the forum.

I can't seem to figure out why my pc keeps failing under load.

I have this pc for about 6 years now. Never had any issues before.

A couple of weeks ago while playing RDR2 (worked fine before) it suddenly crashed, blackscreen, still hearing some stuttering sounds from the game. After a while pc wants to reset itself but can't. It wants to boot but gets stuck with mobo code 43, most of the time, sometimes a different code. If I hold the power button it'll turn off and when I turn it on it's fine again.

Happens after a few minutes of gameplay, sometimes later than other times. So I figured to try 3Dmark timespy and firestrike. Crashes on both at random intervals. HWinfo showed peak temp of 101 Celsius, figured thats pretty hot. So I disassembled my GPU(waterblock) checked for any damage or abnormalities and put new thermalpaste in place. Temp now max 70celsius peak temp, overal about 52 celsius under load.

That didn't change anything. Started really listening to my pc, hear some crackling noise in PSU (RM850i). Figured that might be the problem, shorting out with powerdraw or something. I return the PSU to Corsair, wait for a few weeks. Indeed PSU is faulty. Get a new one under warrenty (RM850x). Install it today, still same problem.

What I've tried so far:

Windows update
Nvidia driver update (also complete clean install removing drivers etc.)
Reset BIOS
BIOS update
Disassembling and reassembling GPU (waterblock)
Reseating my RAM
Trying different PCI slots
New PSU

Specs

OS: Windows 10 64bit
Motherboard: ASrock x470 Taichi Ultimate
Processor: AMD ryzen 7 2700x 3,7ghz, EK waterblock with north and southbridge cover
GPU: Nvidia RTX2080 MSI Sea Hawk waterblock
RAM: Corsair CMR 16gb DDR4 3200mhz
Storage: ADATA SX6000PNP, Samsung SSD 850EVO 250GB, Samsung SSD 870 QVO 1TB
PSU: Corsair RM850X 850watt

I'm out of ideas beside buying a new GPU, but I have a custom watercooling loop so that means a lot of adjustments. Any ideas?

Greetings,
Dutch360
Could be the card is just going bad it happens.
Down clock your card about 15% and try again.