Computer crashing when GPU reaches mid 60 degrees

Status
Not open for further replies.

giles2142

Reputable
Jun 4, 2014
38
0
4,530
So, my issue is that when ever my GPU (HD 6970 powercolor) reaches in the mid 60 degrees the computer will crash, black screen, if there was sound that will loop and it will require and hard reset.

Tried a different card (9800GT). Same issue.
Tried a different PSU (much higher wattage than current) same issue.
Ran memtest overnight, 4 passes no errors.
Tried underclocking the GPU. Same issue.
Reinstalled windows to a seperate HDD, same issue.

Remember that its the same issue with 2 different cards. The rig itself is at least 3 years old, prebuilt from MSY. No issues ever before. Last install of windows at least 2 months ago, nothing updated hardware wise recently, just started out of the blue. Could have been a possible CCC update, not too sure.

IDLE Temp is 50 - 55 degrees Celsius

NEVER Overclocked ANY component ever, only underclocked recently due to this issue, have gone back to stock settings.

Thanks guys! Been at this myself for almost 2 weeks now, im usually quite good with this sort of thing but this has stumped me.

Specs:
PowerColor HD6970
ASUS - P8P67 LE
Antec NeoECO 620C (620W)
16GB Patriot (4 x 4GB)
I7-2600
 
Downloaded RealTemp to watch the CPU temp....... Ran FurMark, GPU got to 90 and CPU got to avg 52 across all core. Seems to have freaking fixed itself... AFTER 2 WEEKS! 🙁 Will report back and continue to monitor if problem presists
 
Well... it happened again... Started playing rust and it died. Same with other games. Can safely say it is not temperature related anymore however it did seem to always crash in the mid 60's
 
The symptoms sounds mostly like a driver issue. Have you tried to roll back the drivers.

Otherwise it kinda sounds like it's a faulty motherboard . What is your hardware info (temps, voltage etc.) Take a screenshot in HWmonitor or something similar and post it here.
 
Tried older driver versions on BOTH installs of windows with no luck. Currently cannot get past the login screen (crahses) to install and screenshot HWmonitor.
Sometimes I get lucky and can get logged on for long enough to get some readings. Like before when I managed to play a game for the first time in 2 weeks for about 20 minutes
 
Without Windows, we can only really look at hardware issues.

Have you got an internal speaker? Does it beep? Does your mobo have AMI or IBM bios (check on line)?

Try a bios/cmos reset and try

Check cpu voltage in bios. Make sure cpu and gpu is not overvolted.

Try another gpu

Try another psu (though the Antec should be good)

Try another cpu and mobo.
 
No internal speaker (that I can hear)
Is a bios/cmos reset safe? Will I lose anything?
Voltage - (http://puu.sh/9eA9f/a0ab05f9c4.jpg)
Another GPU, Same issue (tried a 9800GT)
Tried my mums 850W PSU
GPU Does not crash in my mums comp, (Quad core, 850W, 6GB RAM)
 
The reset will put it all back to defaults and clear any errors

PSU voltages are OK. CPU temp is high if just on idle. CPU voltage is OK

cpu fan rpm?

stock cooler?

blow the cooler out.

If temp still 50C at idle, remove cooler, clean off old thermal paste with isopropyl alcohol, put on new Arctic Silver paste - pea size drop, refit cooler and try again.
 
Got the heatsink off, missing a bit of paste.. seems to have worn out, will replace tomorrow, thanks mate! whats the idle temp meant to be?

btw, the thermalpaste was so worn out, that i could just use toilet paper alone to do it O_O

yes it is the stock fan

Also I clean every component of dust regularly
 
Replaced Thermal Paste with Artic Silver thermal paste.
Still crashing.
http://puu.sh/9fPba/b393648b72.jpg
Reached 50 degrees within minutes just from turning it on and going to the BIOS.

CMOS Battery was out all day while I was at school btw. Still crashing.

Having small issues bootings computer. Sometimes the BIOS wont load up properly on first attempt but will load the next attempt. (Minor issue. Need first issue solved first 😀 )

Unplugging my second screen seems to help, will try playing a game like this and report back.
Running 2 27" Screens, Main is a D2770P 3D monitor. Second is a E2711. Both have been working for at least a year or 2.

Unplugging second screen still crashes, just takes longer
 
Done everything per to that checklist.
Dont have a compatible CPU for my comp to test with so yeah.. motherboard i have tried my PSU and GPU on another board and it works perfectly

Its worth noting that I placed a very crappy GPU just to test with a third card as my motherboard has no on board graphics. The computer did not crash though however I am unable to do much as the card is that bad :/ should have mentioned that earlier.
 
I would like to say that a BIOS update would fix the problem (have you done this?) but it doesn't make much sense as the 9800GT doesn't work either since this a fairly old card. Anyway you should always have the latest BIOS version installed.

I would like to urge you to try one more graphic card that you know work from another system just to completely rule out it isn't faulty GPUs. As I said the 9800GT is an old card so you may just have been really unlucky with the cards you tested.

If you can completely rule out faulty graphics cards and have the latest BIOS version I would say it is definitely a faulty motherboard. When you installed new/old drivers to test, I take it that you did uninstall drivers in safe mode and with driversweeper (example)? Installation of drivers without correctly uninstall old one can easily lead to critical isssues.

It's really hard to say what your problem can be otherwise.
 
The other 2 graphics cards I tried were from my mums and dads computers. the 9800gt has been going for at least 5 years now and is still going without issues in dads computer. The other card in my mums comp (forgot model) has been going for 2 years now as well.

BIOS Update. Same issue.
4 different cards. Only my 6970 and the 9800GT present the problems. A GT210 doesnt have an issue however I cant do anything with it cause it is so shit.. Same with my mums GPU

This is starting to get hopeless.. No one knows whats wrong with it and it all just happened out of the blue. 2 weeks ago I was playing wolfenstein the new order and finished it (had to update CCC i think) and then i moved on to How to Survive and crashes happened during game play of that game. (First thing I tried of course was a system restore to over a month ago)

Both installs of windows. Right after the starting windows screen. just as the log on screen would normally pop up is when the computer is crashing.

2 27" screens on my 6970, = crash (50 Degrees avg)
1 27" screen on my 6970, = can load windows? (30 degrees avg)
Same with the 9800GT, Is this because of my dual screens? How can the screens effect the actual computer? I did not think that could happen?
Never mind...... Just took longer to crash with just 1 screen plugged in... 🙁
 
ffs this is getting really freaking annoying. Plugged the Geforce 210 back in. No crash so far.. but it does with the 6970 and the 9800gt.. only thing in common with those 2 is they need additional power. but i tested the PSU.. anything else I can test before I can test the mobo and CPU with my mates comp in about a weeks time?
 
I think you've done it all -
PSU and CPU voltage and fan rpm OK

You still got same problem with 9800gt.

RAM tested OK with Memtest.

Double check per Tom's checklist - http://www.tomshardware.com/forum/261145-31-perform-ste...

Only thing not tried - new cpu and motherboard?
 
Status
Not open for further replies.