System Randomly Freezes When Gaming

JoshDBoyle

Distinguished
Mar 14, 2011
44
0
18,530
About three or so months ago I built a custom rig particularly for gaming. Here are the specs:

Motherboard: ASUS p8p67 LE
CPU: Intel i5 2500K
GPU: Diamond HD Radeon 6970
RAM: GSkill RipJaws 1333 DDR3
PSU: Thermaltake Black Widow 850W
SSD: Intel x25 40GB (for Windows 7 installation only)
HDD: Western Digital 1TB 6GB/s TR Caviar (Everything else is on this drive)

The issue I'm running into ONLY occurs during intensive gaming (I can have the system sit in Windows for days or weeks and it won't ever lock up and I can watch videos as long as I want without problem). It can happen within the first two minutes of firing up the game or sometimes I'm able to play flawlessly for hours on end until it finally happens. Eventually, however, the system completely locks up (not reboots or shuts down). The image on the screen freezes, the sound stops, the mouse cursor is visible but can't be moved and there is nothing you can do except hard reboot the system.

Most of the time when I hard reboot in this fashion, the system simply takes me to the "Windows did not shut down properly" screen where I select to start Windows normally. Every now and again though, it takes me to an American Megatrends splash screen that states: "Overclocking failed! Please enter setup to configure your system".

I have not overclocked anything nor have I even tried. In fact, I've gone into the bios and made completely sure everything is set to the "optimized defaults".

Here are things that I've tried to date:

1) Originally I thought it was the graphics card given the issue only occurred during gaming so I checked out the temperatures of the card using Catalyst Control Center while running Battlefield: Bad Company 2 (2010) on completely maxed settings. The temperature never got anywhere remotely close to dangerous and was always well within the recommended range for my card.

2) Having ruled out GPU temperature, I grabbed a copy of FurMark and ran it on very high benchmarking settings for about one to two hours. Everything came back fine. Now it's possible that I could have just "gotten lucky" as I said that sometimes I can game for hours before the system freezes so I plan to run FurMark for several more sessions to completely rule out the GPU.

3) After GPU testing and finding nothing I moved to CPU testing and started checking the temperature there. The temps all came back well within the acceptable range without every approaching anything towards any extreme.

4) Next I started looking into memory and realized that the DDR3 memory I had in the rig (it came as a package deal with the GPU from newegg) was not "technically" compatible with my motherboard (by "technically" I mean that it wasn't listed in the mobo manual and I couldn't find it on the QVL list - the RAM, fyi, is Kingston HyperX DDR3 at 1600). So I bought a couple sticks of GSkill RipJaws where I could be sure the exact model number was perfectly compatible. This didn't fix the issue either. I even ran MemTest on both the old RAM (HyperX) and the new RAM (RipJaws) and MemTest came back fine for both. So did the MemOK! setting on my motherboard.


With that all out of the way, I should also mention one thing about the motherboard. When initially building the system, I didn't seat the mobo properly and upon turning my system on for the first time, I immediately smelled some burning and turned the system off. I took everything apart again and took a good look at the motherboard. The only damage I could see was just that a few of the little contact "notches" circling the holes with which you seat the motherboard had been burnt off. Otherwise the board looked fine. I then seated it properly, hooked everything up and all worked perfectly (aside from this random locking issue). Although this certainly COULD be the issue, I'm confused how the system would let me sit in Windows for as long as I want without freezing if the motherboard was fried. Either way, I clearly can't rule out the mobo.



So after all this I feel like I'm left with PSU, Mobo or the hard drives. As a fringe case, a coworker of mine pointed out that it's possible the CPU could be defective in some way but I'm hesitant about this as I can run the system in Windows and have tons of applications open and actively working for hours upon end and the system will never lock up.

Does anyone have any pointers? I'm a bit lost at this point.

Thanks so much in advance.
 

JoshDBoyle

Distinguished
Mar 14, 2011
44
0
18,530
I can be reasonably confident that the GPU is not at fault here. I just ran FurMark again but this time jacked everything way up (8xMSAA, high resolution) and did a burn-in test for an hour straight. No issues. The GPU slowly ramped up to 77 degrees C and stayed there with the fan peaking at about 81%.
 

PreferLinux

Distinguished
Dec 7, 2010
1,023
0
19,460
The only way you can be confident is to try a different card, or to find out it is something else. Furmark may not cause the problem simply because it stresses different parts of the card than when gaming, for all I know.
 

mails2pradeep

Distinguished
Nov 29, 2011
1
0
18,510


I am also facing similar issue of computer freezing as explained by you. Were you able to pin point the reason for this failure. your advice is very much appreciated thank you