Legit Reviews

Whenever I play an intense 3D application my system randomly crashes. It becomes frozen at this screen. It seems to be random, sometimes I can play for hours without problems, sometimes it crashes after 15 minutes. My old video card (6600GT) didn't do this. What do you think is going on here? Bad mobo, video card, not enough power? It only happens in the most graphically demanding games. All clock speeds are stock.

My system:
3800+ X2
Abit AN8 32X
XFX 7900GS 450mhz
Vantec ION2 460W

I got this afterwards.

that would be artifacting, typically caused by overheating possibly caused by over clocking

Everything looks fine until one second it crashes and thats when the artifacts appear. The artifacts appear at the same time it locks up. I thought artifacts gradually appear as the card heats up.

Also, I've checked the temperatures and they are in the 70's celsius under load.

what drivers?

93.71

Should I give the BETA drivers a try?

It could be driver related but first I would google and download drivercleaner pro, then clean the drivers out in safe mode. Then reinstall. See if the problem goes away

Thanks for the help. So far, it seems to be working.

I spoke too soon, it happened again.

My friend was having trouble with stalker, some of the textures were spiking like crazy, there would be random spikes sticking out of areas. Anyway, it ended up being a heat problem, he took off the side of the case and set a fan up next to it and the problem was solved.

Like dicecca said, it may be heat, but not necessarily due to overclocking, maybe you just need better airflow. Try taking off the case panel and playing

Since I got new drivers it crashed twice so far. Both times after less than 30 minutes of gameplay. However, I've also been able to play for hours without crashing. Sometimes I can play for hours without problems. Sometimes it crashes after 15 minutes. How is that consistent with overheating?

I monitored the temperatures, and they don't exceed 80. My case has air holes everywhere, intake, and exhaust fans. Taking off the side panel doesn't even decrease temperatures.

It is not artifacting. There is no artifacts until the crash. At which point the computer locks up. After a few minutes it will display a "no signal" message on the monitor.

70s and 80s are way too high. Its overheating

try putting a case fan next to the video card for a bit with the case door off and see if it helps... It's either heat or a fried memory IC from the looks of it. With the warm summer weather settling across the US it might just be heat though.

Without the side panel the temperature doesn't exceed 78. No problems so far. I might try an aftermarket cooler if it is heat.

Is there a program that can test the memory to make sure its ok?

ya, like already mentioned... it's either heat or bad video memory.

But the last time I saw bad video memory, the screen was garbled even on windows... (Nate should remember our 9700 AIW's ) But your situation could just reveal itself under high video memory usage and the memory IC's used in normal windows and low graphics usage are fine on your card, but one of the last ones used are bad... thus, your card only craps out during high graphics usage.

if you have multiple ram chips does it fill the chip in the front before it moves to the next if you need more ram for what you are doing? Maybe using windows normally it doesnt excede that first chip, but when he starts playing the ram usage requires ram from the second chip, and that is the bad chip.

If you have multiple chips, try putting the chip in the last slot into the first slot and repeat to see if you get any problems.

I just mention this because I have been told that you should always put your best chip in the first slot, then the rest afterward

vicaphit wrote:if you have multiple ram chips does it fill the chip in the front before it moves to the next if you need more ram for what you are doing? Maybe using windows normally it doesnt excede that first chip, but when he starts playing the ram usage requires ram from the second chip, and that is the bad chip.

Great point

Illuminati wrote: (Nate should remember our 9700 AIW's Very Happy )

Yeah how could I forget... ATI sent me three Radeon 9700 AIW's and not one had good video memory on them and I never did end up getting one that worked.

vicaphit wrote:if you have multiple ram chips does it fill the chip in the front before it moves to the next if you need more ram for what you are doing? Maybe using windows normally it doesnt excede that first chip, but when he starts playing the ram usage requires ram from the second chip, and that is the bad chip.

If you have multiple chips, try putting the chip in the last slot into the first slot and repeat to see if you get any problems.

I just mention this because I have been told that you should always put your best chip in the first slot, then the rest afterward

You mean system RAM? I switched the RAM sticks with no results. It is working fine. Also, if something was wrong with the system RAM it would probably fail the power-on self-rest.

How can I definatively test the video card's memory? I don't want to get a new fan and void my warranty if the card has bad memory. I've been trying to get the card to crash, but I can't do it. I've tried maxing out all the settings, but it's not happening.

Have you try running something like 3dmark6 to stress the card? It's not exactly the most controllable test but it will definitely stress your video card and should crash it if that is where the problem lies.

Instead use the artifact tester on atiTool. It will automatically beep when there is an artifact, and can pick up things much subtler than you will generally see. And it tells you the temp of the card as its testing, so you'll be able to see it climb higher and higher.

Dan