My computer crashes fairly regularly when doing CPU intensive tasks. I'm testing it now with a newly downloaded temp monitor and CPU is sitting at 48C under load.
When the machine freezes, I do a manual reboot, and sometimes upon starting back up I get a corrupted screen. I restart and often its fine. I sometimes get the corrupted screen a few minutes after booting back into Windows.
Specs as follows:
Asus A7V880 Mobo
Single 1GB Unbranded Memory (crashes also happened with previous 2 x 256mb memory)
Athlon XP 2800 CPU
Western Digital Caviar 120GB HD
Western Digital Caviar 160GB HD
ATI X800 GPU (with a brand new cooler. Stock cooler suffered same crashes as did previous graphics card)
Vista Ultimate (crashes also happened under XP Pro)
Now the reason I've messed around so much with the graphics card is because the freezing and screen corruption hinted to me that it was a GPU problem. Right now I've got the case open and the GPU fan isn't on, though the heatsinks on the GPU memory are quite warm. I see the fan come on from time to time so it does work.
I'm planning to upgrade most of my components soon but some things I wasn't planning on changing (i.e. Hard Drives, DVD Drives, PSU) can anyone shed any light on what it could be?
For Vista Ultimate you should be running 2GB of RAM. What's the wattage on the PSU? If you can set the GPU fan to continuous, I'd do it. Given the CPU, a 2 generation old mobo, and possibly a weak PSU, you're likely just overstressing the machine. What CPU intensive things are you doing? E.g., rendering/editing?
It sounds as though it is temperature related. Higher termperatures result in greater energy usage and may be outstripping your power supply. A general inspection, cleaning, and reseating of all cards, cpu and heatsink can sometimes work in situations like yours. If that does not accomplish the desired result, you may get by with just a power supply upgrade.
Unless you add a high end vid card that PSU should work fine for a Core2Duo. I would check the voltages in BIOS to see that they are still within 5% for +, and 10% for -. Even quality PSUs wear out eventually.
Your 48c under load is a little high, but not dangerous. It might be time to pull the CPU, clean off the old paste, and reseat with new paste anyway. I did this with my 2600+ after 4 years. Now runs between 30-40c depending on load. I used Arctic Ceramique.
The freezes definitely seem to be vid card related. I have never been able to get one of the newer cards to run on my 2600+ system. I am still running a Sapphire 9600+ 256mb on the Shuttle Nforce2 MB. I have tried Nvidia 6200(lots of flicker), and a Power Color x1600 pro 512. Got all kind of weird screens with the 1600, but no lock ups.
I tried over volting the AGP, lowering ram timings, even under clocked the CPU. Nothing including several versions of drivers would make those other cards work. Finally reverted back to the 9600, and the driver it came with. No more issues.