Graphics card failure?

My computer has the following symptoms with no warnings (gadgets included);
Works fine for an indefinite amount of time.
Computer freezes (on the rare occasion quickly judders to a halt then freezes).
Screen goes black (monitor in standby mode).
Sound stops.
Computer shuts down and restarts.
On start up I'm told it didn't shut down properly and after windows has booted up I get this error;

********************************
SourceWindows 7 Pro 64bit

Summary
Shut down unexpectedly

Date
‎23/‎07/‎2013 18:13

Status
Report sent

Problem signature
Problem Event Name: BlueScreen
OS Version: 6.1.7601.2.1.0.256.48
Locale ID: 2057

Extra information about the problem
BCCode: 116
BCP1: FFFFFA8005ED74E0
BCP2: FFFFF880057ADE30
BCP3: 0000000000000000
BCP4: 0000000000000002
OS Version: 6_1_7601
Service Pack: 1_0
Product: 256_1
Bucket ID: X64_0x116_TdrBCR:2_Tdr:2_IMAGE_nvlddmkm.sys
Server information: 5eedc44d-878f-4a0b-afef-b6fed4b1aaf4
********************************


Backstory:
I've had a custom built computer (but not overclocked etc in any way) from 2009 mostly with a 4870x2 (+custom cooler) and it's worked for a long time but must have started having problems as I swapped it with a 8800. This then succumbed to the same problem after a while (but it was from an older computer which could have damaged it so won't rely on it too much). It used to happen when I was playing a game then start playing Windows Media Player so used itunes. About a month ago the problem became so bad just using the internet caused it to crash or loading a game but on other occasions I could have 3 versions of the game playing no problem, there never seemed to be a pattern. It then happened every time I tried to load a game.

In the end all the evidence pointed to the GPU so bought a new cheap stock 620 on a clean computer. This has worked fine for 1 month but that too is now showing the same problems although not as frequently it seems. I've got gadgets which check my CPU and only once did it warn me it had hit 82 degC then shut down (this rarely happens). My GPU has been no higher than 69 degC and when my computer's crashed the last few times it's been around 56 degC.

I've now had it crash twice in close succession whereby I had 3 copies of a game running in low settings (2nd time it happened the 2 copies I wasn't active in were CPU limited and was ~60 degC). I alt tabbed between all 3 game copies to try and test the load and after a few minutes it crashed at 56 degC. The GPU fan speed was constant at 65%, load was 99% and memory used no higher than 700Mb/1Gb.

I'm not entirely sure the GPU is the culprit and could be the RAM or motherboard (one place suggest maybe even the sound card but I don't know if they were talking integrated which mine isn't). Reading around suggests it can't be the CPU because my computer wouldn't work at all. What checks I could think of show no problems and given that my computer worked fine for 1 month after terminal problems suggests it is the card and I'm just unlucky or expecting too much from a cheap card especially given the prompt failures under testing.

I'd appreciate any help although I'm aware that this sort of problem is particularly hard to solve and spent a lot of time trying to resolve it myself. I'll do my best to answer any further questions but my expertise is probably limited.

P.S. I'll uninstall and reinstall the driver again as I did update the driver a week or so ago to the latest driver using geforce experience but didn't uninstall the previous driver (thought nvidia did that for me). Doubt this will fix the problem but worth a shot.
 
Windows Reliability monitor gives this information each time I encounter a crash (typically it logs it twice followed by a log of shutting down unexpectedly).

Source
Windows

Summary
Video hardware error

Date
‎23/‎07/‎2013 18:13

Status
Not reported

Description
A problem with your video hardware caused Windows to stop working correctly.

Problem signature
Problem Event Name: LiveKernelEvent
OS Version: 6.1.7601.2.1.0.256.48
Locale ID: 2057

Files that help describe the problem
WD-20130723-1811.dmp
sysdata.xml
WERInternalMetadata.xml

Extra information about the problem
BCCode: 117
BCP1: FFFFFA800708A010
BCP2: FFFFF880057ADE30
BCP3: 0000000000000000
BCP4: 0000000000000000
OS Version: 6_1_7601
Service Pack: 1_0
Product: 256_1
 
Both those errors are directly caused by video drivers. First, have you checked to see if there is any dust built-up in your system and particularly your graphics card.

Second, The VIDEO_TDR_TIMEOUT_DETECTED bug check has a value of 0x00000117. This indicates that the display driver failed to respond in a timely fashion.
 
The current 620 GPU has no dust at all as it's only been installed for 1 month. The motherboard itself is relatively dust free with dust only around places like the fans (CPU, chassis floor and aft and front hdd).

I've uninstalled the driver and installed from scratch.
CPU temps have been 41-74 degC
GPU has been 36-68 degC.

I'll check my computer interior thoroughly tomorrow as well as repeat the same loading tests I did earlier and let you know the results.

Thank you.
 
74C = 165F
68C = 154.4 F

Check your recommended temperature specs for your gpu and cpu.
 
According to Nvidia the max temperature is 95C which the card has never reached to my knowledge. The only accumulation of dust I saw was on the top of my sound card and hdd which I cleaned.

Since my last post 3 days ago my computer hasn't crashed at all until today.

Earlier it handled 3 copies of a game no problem. Later though it crashed just having one and just now, two copies. The graphics temperature is in the mid 50'sC when it crashes and my CPU I'm not sure but I've had no warnings to tell me it's too high. I've seen one of my 4 core hit 71C but it didn't crash.

I'm wondering if I should next remove my fans and give them a thorough clean. What are the chances of this being a CPU,RAM, motherboard or PSU issue whether it be hardware or heat related?
 
Let's first try a simple yet tried and true diagnostic step. Open the side of your computer. Now take a small fan and on the low setting have it blow directly into your case. Run the PC as before, I.e., games,etc. that caused your crashes.

I realize that this may take a day or two but let it run like this as long as maybe 3 days. What are your results?
 
Hello again.

After a good few days of having a desk fan blowing in to my computer when I was asking it to run 3 games at once and suffering 'no' problems I stopped using the fan. For a week or so now I've been able to run everything I normally would without the fan without problems.

Someone suggested maybe I blew some dust loose though I can't see what dust could have caused the problems I had. My CPU fan is fastened together by allen screws but lack the correct sized wrench to take the fan out to clean it properly.

Until my problem comes back and persists I'll assume my computer is working properly and consider this thread closed. Thank you again for your help.
 
Back