Computer keeps freezing while gaming, usually after 20 minutes or so. No BSOD. Doesn't react to any commands

[Mental Allergen]

My specs are as follows:
MOBO: Gigabyte GA-990FXA-D3
CPU: AMD FX-4100
GPU: AMD Radeon HD 6950
PSU: Cooler Master GX Bronze 750 W
Using W7 64x

The freezing occurs always while gaming, neither while browsing or watching movies. I'm suspecting a software issue but the problem might be caused by malfunctioning hardware, after all the machine is at least 2 and a half years old
I'd gladly appreciate any help on the topic, I'm getting really annoyed when my system freezes while playing kerbal space station!

 

[IamTimTech]

I would try your power supply first.

 

[RaDiKaL_]

Monitor your temps, games being very demanding it's possible that your CPU/GPU are overheating leading to the freezes you're suffering. Monitor them with HWMonitorPro or AMD Overdrive (download it from amd website) and GPU-Z for the video card.

 

[Mental Allergen]

The system isn't overclocked in any way.

Just tested temperatures using HW Monitor and GPU-Z. On idle they are around 30°C for the CPU and 41.5°C for the graphic card.
Did a prime95 stress test on the CPU, approximately 5 mins and force ended it when the CPU got to 74°C.
Did the Furmark GPU stress test and 8 minutes and 24 seconds in it froze, had the reset the computer. I had been running steadily for more than 5 mins until it froze with GPU on 74°C on a configuration as follows: 1600*900 fullscreened with 2x anti-aliasing. Any opinions on these results?

 

[RaDiKaL_]

Sounds like your GPU is the issue, surely it got to an even higher temp level which caused the freezing and never got logged, your CPU temp during load isn't good either, it shouldn't go over 70ºC even with an intensive test like Prime95.

Do you regularly clean the dust from your computer? like every 3-6 months at most? maybe it's time to give it a good cleaning, make sure fans are spinning with no problems, might be time to replace the Thermal paste in your CPU (usually lasts 1 1/2, 2 years).

 

[Mental Allergen]

Sounds like your GPU is the issue, surely it got to an even higher temp level which caused the freezing and never got logged, your CPU temp during load isn't good either, it shouldn't go over 70ºC even with an intensive test like Prime95.

Do you regularly clean the dust from your computer? like every 3-6 months at most? maybe it's time to give it a good cleaning, make sure fans are spinning with no problems, might be time to replace the Thermal paste in your CPU (usually lasts 1 1/2, 2 years).

Thanks for the input.
I think it's been well over half a year I've cleaned the machine from dust, will do it again soon.
I've already checked the fans and they seem to be running ok but maybe I could amp them up a bit manually? Furmark said the fan was doing 60% or so while the test ran. Is that low or not?
About the thermal paste, I admit, I haven't a clue about how to do that or where to buy one or even how to apply it, but google's gonna be my brother on that.
I'll do the things you mentioned and then I'll report again here.

 

[IamTimTech]

I don't think your gpu or your temps are the problem, but to test it try using your computer with the GPU removed using your dedicated graphics on the motherboard or a different gpu all together.

If your gpu were the problem it would A: eventually unfreeze B: probably give another indication that it is failing such as artifacts and/or loud fans.

A bad power supply can cause a system to hang up and then not recover. Swapping out pieces one by one till the problem goes away is the only sure fire way to find the culprit. Just taking out the GPU and using the dedicated graphics won't necessarily tell you what the problem is because if you take out the graphics card you may not be drawing enough current to make your power supply faulter, and/or you won't know if the card itself was the cause. To test the card itself stick it in a friends machine for a while.

I normally would be inclined to blame the gpu before a psu, but cooler master isn't the best PSU manufacturer (Not the worst by far) and those symptoms describe a bad psu very well.

my 2 cents.

 

[Mental Allergen]

A bunch of things happened since the last time I posted an update, so now I'll list them, but first, I gotta say I did some work on the computer.

I took out the processor and the heatsink, cleaned it up and had it installed again. One note here is that I cleaned the excess thermal paste there was and since I haven't got any, I didn't put it. Yeah... There is no thermal paste between my processor and the heatsink.
Also, I vacuumed up the housing, cleaned the graphics card a bit and the ventilators. That's it.

Two days ago, again while gaming, twice it has happened that the machine simply shut down. No warnings, just dead. Somewhat related to that, I checked the windows event viewer and under the system tab i found a bunch of kernel critical warnings listed. They were all events with ID 41 saying that the system had rebooted without cleanly shutting down first. So far, I'm blaming the PSU.

Soon, I'll have the computer sent for a proper dust cleaning, and also I'll be getting some of that thermal paste to apply it and a new power cable because mine's really old. Frankly speaking, I don't feel a bit comfortable without the thermal paste but the processor managed to level at 60 degrees C while the prime95 ran for about 10 min.

 

[IamTimTech]

There is no thermal paste between my processor and the heatsink

You must have thermal paste, this is not optional. Go to your local radioshack or best buy and get some generic silicone paste at least. Your computer will continue to overheat until you do this.

 

[Mental Allergen]

You must have thermal paste, this is not optional. Go to your local radioshack or best buy and get some generic silicone paste at least. Your computer will continue to overheat until you do this.

I know it isn't optional but I firmly believe that this isn't and overheating issue. All the crashes I had happened during low loads of CPU and GPU. The CPU was doing about 40 degrees while the GPU was at 45-6 while it crashed last time. I've got a nice airflow going through and a cool room.

 

[IamTimTech]

Then you and I agree, it is the PSU. None the less get some thermal compound on that rig my friend.

 

[Mental Allergen]

Alright, I'm back with a few interesting happenings that occurred.

I had the machine dust cleaned with a proper compressor. I've applied some thermal paste. Now the processor idles at 18 degrees and doesn't go over 40 degrees no matter what on the prime95 stress test. Temperatures on the graphic card have stayed the same (idling at 40 degrees). Weirdly enough the crashes and freezes have stopped happening. Huh. Maybe it was an overheating issue in the end. D:

I haven't changed the PSU or the cable, which I had in mind to do, but now it seems there ain't any need to do so. I hope that the problems won't surface again

 

[IamTimTech]

That's awesome man, congrats

 

[slyu9213]

I'm just glad your CPU didn't bust for not having any thermal paste for a while. I had crashing issues with one of my PC rigs when playing games and it really frustrated me too. I still haven't made a sure conclusion but I learned one that my RAM can't run at it's rated speeds (too much errors) so I have it underclocked currently and my GPU is also not as strong as it use to be. Glad you have no more crashes.

 

[Mental Allergen]

I had crashing issues with one of my PC rigs when playing games and it really frustrated me too. I still haven't made a sure conclusion but I learned one that my RAM can't run at it's rated speeds (too much errors) so I have it underclocked currently and my GPU is also not as strong as it use to be.

That's actually very interesting to me because I've got two pairs of Kingston HyperX Silver ram's and their native frequency is 1600 mhz but my motherboard can only accept up to 1333 mhz. None the less, i haven't had any problems with the ram, it even passed an all-nighter on memtest86+

Ram isn't expensive, you could try with new one, maybe that would speed up your machine

 

[Mental Allergen]

I had crashing issues with one of my PC rigs when playing games and it really frustrated me too. I still haven't made a sure conclusion but I learned one that my RAM can't run at it's rated speeds (too much errors) so I have it underclocked currently and my GPU is also not as strong as it use to be.

That's actually very interesting to me because I've got two pairs of Kingston HyperX Silver ram's and their native frequency is 1600 mhz but my motherboard can only accept up to 1333 mhz. None the less, i haven't had any problems with the ram, it even passed an all-nighter on memtest86+

Ram isn't expensive, you could try with new one, maybe that would speed up your machine

 

[aqe040466]

Definitely the Power Supply Unit(PSU) Change it.

 

[slyu9213]

I had crashing issues with one of my PC rigs when playing games and it really frustrated me too. I still haven't made a sure conclusion but I learned one that my RAM can't run at it's rated speeds (too much errors) so I have it underclocked currently and my GPU is also not as strong as it use to be.

That's actually very interesting to me because I've got two pairs of Kingston HyperX Silver ram's and their native frequency is 1600 mhz but my motherboard can only accept up to 1333 mhz. None the less, i haven't had any problems with the ram, it even passed an all-nighter on memtest86+

Ram isn't expensive, you could try with new one, maybe that would speed up your machine

Ha, well the motherboard I was using officially supports the 2400MHz speed. I just think I got bad modules. But I bought these for $64 so I think it's pretty cheap for being CAS 10 8GB 2400MHz RAM. GSKill has Lifetime Warranty so I'll just RMA them when they get any worse.