I built a computer several months ago, and ever since then I've been having problems with crashes during games, and even more recently, random restarts. The crashes seem to be random; they'll happen 15 minutes into a game or 3 hours or sometimes not at all. I rarely go over 60% RAM usage, and the video card temp stays right around 70-75 degrees C.

I have:
AMD Athlon 64 X2 Dual Core 3800+
Biostar TForce4 SLI (NF4 SLI)
nVidia GeForce 6800XT (256MB RAM)
G.Skill F1-3200PHU2 1GBx2
Rosewill RP550-2 PSU
Soundblaster Audigy (Model SB0090)

I have all the latest drivers for all of the hardware.

My boyfriend and I (who has been helping me out with this problem) think the problem may lie with the memory, graphics card, or in worst case, the motherboard. But, we haven't been able to pinpoint anything exactly. I've looked in forums everywhere and other people seem to have similar problems with the memory and even more people with the graphics card... so I don't know. Any advice would be greatly apprectiated.

Thanks!

My guess would be the power supply. Rosewill is not a manufacturer, they're a repackager. They buy products made by others, then slap a Rosewill label on them. That model is probably made by ATNG POWER CO LTD & although the specs look reasonable, the fact that there's 50A on the +5v rail is very suspicious. Very few components use the +5v rail so 20A or so is generally sufficient...the 30A extra are completely useless, but they do manage to add 150W to the PSU wattage rating. This is more closely a 400W PSU than a 550W PSU. I had a look at newegg & I see that it gets good "customer reviews", but most that I read mentioned how cool it looks...big deal!

Before running out & buying anything, make sure to troubleshoot. Random reboots & crashing are classic sysmptoms of a weak PSU, but there are other possibilities. What is your CPU temp? Have you tested the RAM using memtest86? Are all you drivers (especially motherboard chipset & video) the most recent available? Are you sure you BIOS is setup correctly? Do not simply choose "setup defaults" or "optimized defaults". The BIOS should always be manually configured to best match your hardware.

You said it has been like that from day one, I would suggest you clean out all the fans in the system in addition to dismantling the CPU/HSF.

Make sure the CPU has the right amount of thermal film (thin layer) after you clean off the old coat and also ensure proper contact between it and the HSF when you reassemble them back.

I'm not up on video card temperatures, but 70-75C seems pretty hot to me. I'm curious about the CPU and especially the 'case' temperatures.

Do yourself a favor BACKUP!

Hi

Your problem may not be the hardware at all. I had the same issue and I replaced nearly every component, then as a last measure I reinstalled the OS and the problem went away (microsoft).

Lez

I did run memtest, with no errors, and I tried testing each stick individually with various games and I got crashes both times. The BIOS is fine, and my CPU temps idle right under 40 degrees C.

I downloaded Windows Debugging Tools and used dumpchk.exe to read one of my latest dump files... it told me that 'e10kx2k.sys' was most likely the cause, which, upon googling, is a driver for my sound card. Guess I had kind of forgotten about updating my sound card drivers; they're a couple of years old. So I'm downloading the new drivers now and hopefully that should fix the problem. And that actually makes sense, because sometimes the sound will randomly cut out in games and when I exit the system crashes (sometimes).

And about the PSU, it's rated for 30a, not 50a on the +5 rail. It has two 12V rails at 18A each, which is what makes it 550 watts and not 400. I know that it's not the greatest PSU ever... but I really don't think that's the problem.