This is a self built PC and has been working well since Christmas. I had issues with the RAM where one stick was faulty and had a new set which I installed a week ago.
Computer was running fine for a few days, being on 10 hours a day, then I had some issues whilst playing WoW. First the sound seemed to go off and come on, then I got graphical artifacts. Eventually the computer started to freeze completely causing me to force restart it even outside of any application (ie at idle). I assumed this may be due to heat so reseated everything and applied new thermal paste to the processor. Whilst doing this some of the pins got caught by the cloth inside the socket which moved them into awkward positions. I used twisers to move them back into position and eventually got the PC to boot again.
However the original problem is still there where the PC will freeze and hang there. Sometimes the image remains, sometimes the monitor will go black (the graphics card has turned off i believe). Occasionally I will get a BSOD with a random BCCode.
If I run in safe mode the PC works fine, occasionally the PC will boot as if it has no video drivers and appears to be stable like that. I have tried installing the latest drivers whiched stopped the artifacts but did not stop the freezing.
To add to any confusion DXdiag previously showed the full 6GB of RAM and now only shows 4GB. I have tested each stick on it's own and they have all POSTed.
I am unsure what I can try next. I'm worried I have damaged my motherboard's socket enough that I will have to replace it although the issues I'm having occured before this. I have not yet reinstalled windows, is this likely to help?
(HT was disabled stressed with Prime95, read with SpeedFan)
I'm reinstalling windows as we speak so will get some more when I can. But GPU was idle at 50-60 Degrees C and only got up to around 70-75 Degrees C max when I viewed it. Think under load it used to get to 90 Degrees C which I've been assured is normal by the manufacturers.
Good god.... 90*C, and that is normal? I run 2 9800GTX+'s and a GTS250 (3-way SLI), when they reach 70*C they start to artifact.... Download EVGA's precision utility and crank the fan up to 100-75% and that should drop your GPU temp, and fix the artifacting.
90 Degrees was using FurMark to 100% stress the GPU. Under normal load (3D Mark) the temperature is much more like 80 Degrees. The throttle for the card I have is at 105 Degrees, I had an old card that approched that under FurMark which I had replaced. The GX2 is 2 cards slapped together so the elevated heat is expected.
Also as I mentioned there arn't any artifacts now, the issue is the freezing, seemingly when graphics drivers are installed. The temperature has only been to 75 Degrees at most that I've noticed with these issues.
I have just done a clean install of Windows 7, Updated the BIOS, Updated various drivers and am now trying it with the nVidia latest drivers for my card. If that holds I'll do a temperature test on it to get a more reliable reading.
I've been trying to install drivers for the graphics all evening but it wont accept any but the ones that came on the disk. Downloading and installing the latest drivers is successful but on restart there are no drivers.
Reading around it seems I have a common problem, "Windows has stopped this device because it has reported problems (Code 43)" showing under one of the graphics cores. SpeedFan only reports on GPU where it used to show two so looks like there is a hardware fault on the card / maybe the damage to the pins on the motherboard socket?
I believe the card is still under warrenty so I will contact them tomorrow and see what they say.
To update where I'm at now. I contacted my card manufacturer (PNY) and they have said they will RMA the card and upgrade me, as I've sent one back before, to a GTX 470, which seems like a generous upgrade from what I've been reading.
That hopefully solves the graphics issues but it still remains that my PC wont recognise more than 4GB of RAM. Each stick seems to run ok, will try and do a memtest or something similar to confirm. The BIOS also only reports 4GB of RAM being installed when all are slotted. Would this suggest an issue with the motherboard?
The PC seems a lot more stable now windows has disabled one of the GPUs. I ran FurMark for 20 Mins with each stick of RAM and then all together.
Temperatures were 50-60 Degrees Idle and capped at 84 Degrees during FurMark.