[SOLVED] Artifacting under load leads to crash

[shadeshot01]

i have a 1080ti ftw3 with a kraken g12 + kraken x72 cooler on it with no extra heatsinks for the vram or pwr, when i do a "light" test, like the one in occt, my gpu will artifact then crash in about 30 seconds, if i have a small fan meant for a room pointed at my gpu it can go longer than 10 minutes, with that i thought it was a temperature issue and ordered some heatsinks to put cool the smaller parts
i did more testing to see what games im able to play without my gpu crashing and noticed that when starting games my gpu will crash even with lower temperature reading than my occt test but the power spikes to 280-290w and settles around 220w, during occt it would spike to 240w and fluctuate at around 180w
included (hopefully) are 2 pictures that show precision x1 on the icx temperature tab as well as hwinfo65 when i launched a game and hit the main menu (risk of rain 2), took from a short video i filmed to catch the readings right as it freaks out

i didnt really monitor my gpu temps with the stock cooler but the temps im getting seem low to me when i research what other people were reporting as high temps for vram and power and i had no issues even with an overclock on the gpu core and memory
hwinfo and precision x1 freak out right before the crashes so it might juts be that the readings dont update fast enough

extra note, when i take the side panel off and put it back on my pc it rubs against the power into the gpu and that causes the same artifacting and crash, worst case my gpu is messed up and i need a new one, 2 better cases is the heatsinks im getting fix it or i need a new psu, my current one is 5-6 years old
i also dont have another pc or any parts to switch around to test the gpu or psu

edit: another thought, i read somewhere that cpu thermal paste could cause problems if it leaks off the gpu die, saying that some pastes are captive (i think is the word they used) and will cause problems if it comes in contact with anything other than the gpu die itself, when i take the cooler off to put the extra heatsinks on i will try to clean up anything that might have leaked off the die, will update this post if there is improvement after cleaning + heatsinks

vv specs vv
MOBO: ASUS ROG MAXIMUM XII HERO (Wi-Fi)
CPU: i9-10900k @5.2ghz
GPU: EVGA GTX 1080Ti FTW3
RAM: 32gb TridentZ G.Skill 3200mhz
PSU: Corair CX750M V2 80 Plus BRONZE

https://imgur.com/a/VWEzgi2

View: https://imgur.com/a/VWEzgi2

 

[vov4ik_il]

Artifacts are usually caused by overclocking and/or memory and/or memory controller issues.
GPU memory can be replaced at a reasonable cost, the controller is integrated into the main chip.
I would first remove overclock if any (or even downclock) and test. If it yields the same result - repair or sell for parts.