GPU stops sending signal (dies), fan revs to 100% yet system still fully functions. HELP!!!

Chris Wallace

Reputable
Mar 12, 2015
4
0
4,510
Hey any help is appreciated. It's like no problem I've encountered before. Ever since Tuesday my GPU has been crashing and the fan revs to 100% so I just sit with a black screen. I know the system still functions normally because I was in skype with my gf one time when it happened and I was able to tell her I had to be right back. It happens under load, or at idle, temps are between 43c-85c that I have seen. So far if I have speedfan running and crank the fan up to like 70% it hasn't crashed yet. But it's so spontaneous I'm not sure if that's a sign or just a fluke. It can go most of the day without crashing and then poof it wont even let me on windows for 2 mins without crashing again. Seems to work better after I let stay off for a while. The voltage reads normal for 12v, 5v, 3v, etc. Not sure if it's powerspikes for random heat bursts in the GPU. Nothing is overclocked. I've had this build for like 4-5 years and it's never had a problem. Thanks <3

Specs:
i5 2500k cpu
xfx radeon 6970 gpu
cougar 700w psu
asus sabertooth p67 mobo
8g gskill ram
 

Chris Wallace

Reputable
Mar 12, 2015
4
0
4,510


Kernel-Power 41 is the only critical logged, but I have 0 clue wth that means. There are some misc warnings and errors too is you want those
 

Chris Wallace

Reputable
Mar 12, 2015
4
0
4,510


Also, I just saw that hard rebooting causes that error to be logged as well, in which I've had to do every time the GPU crashes.
 

Mysticking32

Honorable
Sep 28, 2014
548
1
11,365
That error usually refers to an error with the psu.

But more than likely it's probably something more simple. Try reinstalling your graphics card drivers and ensuring it's the latest version.

(And no don't worry I don't need the other errors)
 

Chris Wallace

Reputable
Mar 12, 2015
4
0
4,510


I installed the newest drivers last night and it crashed this morning. However, I updated my mobo bios and (fingers crossed) it hasnt crashed in 5 hours.

Not to be overthinking this, but I'm stress testing my gpu right now with furmark and its at 95-100C is that too high? I'm wondering if the GPU thermal paste is getting old.