Games Crash to Desktop

[DChuv]

Hi, guys.

Yesterday I replaced my CPU cooler (the old one started failing), and now any video game that I run crashes to desktop. The less demanding ones crash after an hour or so, the more advanced ones -- after a few minutes.

My CPU temps have improved substantially: the thermal margin fluctuates between 15 and 30 C (while it used to stay within the 0-20 C range). So, I have a feeling that this new problem is somehow related to the GPU voltage. Yet, I am not knowledgeable enough to resolve it. Can someone advise, please.

Thank you in advance.

Sincerely,
Denis

Here are my specs:

AMD FX-8350
Asus M5A97
GeForce GTX970
16 Gb Ram
Thermaltake 750W PSU
Raidmax Smilodon Extreme ATX-612Web

 

[qwebbq]

I doubt it's GPU, but safe voltage for a non clocked 970 should be around +/- 0.6 V for 12 rail and ~1.35 V or below for core/VRM under heavy load.

 

[Karadjgne]

Check Windows Event Viewer. It has a listing of all the errors that happen, from little to critical. Do so exactly right after a crash, so you'll get an accurate look.

Games crashing to desktop are usually caused by a gpu driver issue, not voltage. Possibly it might be temp related on the gpu, but this can be monitored too. If you have the latest drivers for the gpu, its possible you got a 'blurp' in the dl that's creating an error now. I'd go to Guru3d.com and download the DDU, a program that'll fully delete the current video drivers. Then install a fresh copy previously saved. Don't use the original ones.

 

[DChuv]

Check Windows Event Viewer. It has a listing of all the errors that happen, from little to critical. Do so exactly right after a crash, so you'll get an accurate look.

Games crashing to desktop are usually caused by a gpu driver issue, not voltage. Possibly it might be temp related on the gpu, but this can be monitored too. If you have the latest drivers for the gpu, its possible you got a 'blurp' in the dl that's creating an error now. I'd go to Guru3d.com and download the DDU, a program that'll fully delete the current video drivers. Then install a fresh copy previously saved. Don't use the original ones.

I checked the Event Viewer. It does indicate that some errors took place (exception codes 0xc00000fd and 0xc0000005).

Uninstalled the drivers and then reinstalled the latest version. It did not help.

Monitored the temps of the GPU: it stays in the 60's and lower 70's C. During the last crash the GPU temp was 59 C. So, it appears that the temperatures are fine.

I am no expert, but it does seem to me that this is a hardware-related problem. Not counting the heating issues, everything was running perfectly well until I replaced the CPU cooler. I did not install, uninstall, or update any software. Thus, I think it is unlikely that some software suddenly decided to act up.

 

[Karadjgne]

Heh. You run Windows. Guaranteed there were updates you knew nothing about, unless you have windows pro or Enterprise, in which case you can stop auto updates.
Both of those errors point towards a corruption somewhere, and quite often ram related as software/drivers are accessing addresses that they aren't supposed to.
System File Checker is a handy tool included with Windows that allows you scan for and restore corruptions in Windows system files (including those related to Error 0xC00000FD).
To run System File Checker (Windows XP, Vista, 7, 8, and 10):
Click the Start button.
Type "command" in the search box... DO NOT hit ENTER yet!
While holding CTRL-Shift on your keyboard, hit ENTER.
You will be prompted with a permission dialog box.
Click Yes.
A black box will open with a blinking cursor.
Type "sfc /scannow" and hit ENTER.
System File Checker will begin scanning for Error 0xC00000FD and other system file problems (be patient - the system scan may take a while).
Follow the on-screen commands.
Run Windows memory diagnostic
Run memtest overnight.

I'd also go into Windows update settings and see if it'll let you choose when auto updates will install. Set for idle times after 15mins or so.