Games crashing repeatedly

Laurih · Dec 10, 2013

OS: Windows 7 Home Premium

Hardware:

MB: ROG Asus Crosshair V Formula (Bios version 1703)
CPU: AMD FX-8150 (Overclocked to 4,2GHz)
GPU: XFX Radeon HD R7950 (No overclock)
PSU: Nexus RX-1K (1000W)
RAM: 4x 4Gb Kingston HyperX 1600Mhz

Games crash after couple of minutes causing the computer freeze. Usually I can Alt+Ctrl+Del to Program manager where it tells that game has crashed. My browser also crashes at the same time.

Couple of days back screen went to power safe mode after the freeze and stayed in it unless I switched the power off from the PSU. I cleaned the computer and it seemed to stop complete crash.

This seems to happen only when GPU reaches over 75 Celsius. This should be the normal temperature during stress (Checked it from the manufacturer).

I have also done Memtest and no problems were found.

ihog · Dec 10, 2013

Does this happen when your CPU isn't overclocked?

Skeefers · Dec 10, 2013

AMD recommends that their FX series CPUs don't go over 62C; 75C is WAY too high, and is most likely the cause of your crashes. I suggest removing you overclock until you get a better cooler for your CPU that can handle the higher temps generated by overclocking. If you have a good cooler (what are you using, btw?), perhaps your thermal paste is either inferior or not applied correctly.

I run an FX-8350 on the same motherboard you have, overclocked to 4.3GHz with an Antec Kuhler H2O 920 cooler, and my temps at max during a stress test never pass 35C; you should be aiming to have your temps in that same range.

Edit: I now see that you said your GPU temps are at 75C, not your CPU temps; my apologies. Perhaps your CPU isn't receiving enough power for the overclocked speeds you're using...

ihog · Dec 10, 2013

Skeefers :

75C is the GPU temp, not the CPU temp.

EDIT: Sorry, just saw your edit.

Laurih · Dec 10, 2013

I tried with no OC but the program+browser crashed excatly when GPU at 75 Celsius.

expl0itfinder · Dec 10, 2013

That is strange, as 75C should be just fine for a GPU. Have you tried doing a clean driver install? You may have corrupted yours.

Laurih · Dec 10, 2013

Laurih :

expl0itfinder :

Clean driver install done, didn´t help. It seems that 75 C is the magical point were programs start crashing, because after i turn my chassis fan speed higher it continued normally (Temp was 65C after that).

ihog · Dec 10, 2013

That's odd. 75C is practically nothing for a GPU. Could you also monitor your graphics card's VRM temps? HWiNFO can do that.

Laurih · Dec 10, 2013

ihog :

VRM temp 1-2: Idle: 30C-33C Just before program crash: 65C-69C

I also got a notification: Display driver stopped responding and has recovered.

gokul36 · Dec 10, 2013

Try changing the CPU and GPU coolers

expl0itfinder · Dec 10, 2013

Well then it likely isn't a heat issue. If it were heat, it would probably crash the entire system, not just cause a temporary driver crash. What driver are you using atm?

Laurih · Dec 11, 2013

AMD Catalyst Control Center 13.9. I did a clean driver install but that didn´t help. Seems like software problem, since everything is working normally even at 74C, but after that screen shuts down, and all programs crash.

gokul36 · Dec 11, 2013

Backup ur system....and try to Install a Clean Copy of OS

Windows 8 Pro is much better

expl0itfinder · Dec 11, 2013

You don't need windows 8 pro, but a standard copy of Windows 8.1 can't hurt. Let's save the OS restore until last, as that is a hassle that we want to avoid if possible. Let's narrow the issue a bit. Have you tried running prime95, or Furmark?

Prime95, to stress the CPU: http://www.mersenne.org/freesoft/

and Furmark, to stress the GPU: http://www.ozone3d.net/benchmarks/fur/

In prime95, you will see an error saying "WORKER STOPPED" in one of the boxes if the CPU is instable. In Furmark, the system should crash altogether if the GPU is unstable.

Laurih · Dec 11, 2013

Prime95 reported no problems, but Furmark caused a system crash at 76 C similar to the crashes I encountered week ago.

Two fans at my GPU worked fine at 50% at the time of the crash.

expl0itfinder · Dec 11, 2013

Laurih :

Try setting a custom fan curve, and having the GPU fan/s run at 100%. It will be loud, but we want to eliminate the possibility of a bad temperature sensor. If the GPU still hits 75C under 100% fan curve, and crashes, you probably have a faulty sensor.

Laurih · Dec 11, 2013

GPU wasn´t able to hit more than 68C with 100% fan speed at Furmark. No problems that I could see at that temperature.

expl0itfinder · Dec 11, 2013

Did it crash?

Laurih · Dec 11, 2013

No crash.

gokul36 · Dec 11, 2013

I had the same Problem....before 1 yr back....

I cleaned my GPU fans and applied new Thermal paste..... and now i just work fine....

expl0itfinder · Dec 11, 2013

Yep, then it is merely a heat issue, nothing too special. As the poster above me said, clean those fans, and some Arctic Silver 5 couldn't hurt anything.

Laurih · Dec 11, 2013

I cleaned my GPU fans 2 days ago and that didn´t help. Also CPU temperatures are relatively normal.

I´ll clean GPU fans again just in case.

expl0itfinder · Dec 11, 2013

I'm talking about on the GPU. GPU's have TIM just like CPU's do. Sometimes, the TIM wears, and becomes less effective. The TIM on your GPU may be working as more of an insulator than a conductor. If you clean and reapply the TIM on the GPU, you should see some significant improvement.

Laurih · Dec 11, 2013

Oh GPU. I´ll try and inform if there is change!

gokul36 · Dec 11, 2013

@Laurih do u stil have warranty covered for ur Product?

Games crashing repeatedly

Honorable

Distinguished

Honorable

Distinguished

Honorable

Honorable

Honorable

Distinguished

Honorable

Distinguished

Honorable

Honorable

Distinguished

Honorable

Honorable

Honorable

Honorable

Honorable

Honorable

Distinguished

Honorable

Honorable

Honorable

Honorable

Distinguished

Share this page