Artifacting and crash

Today I started having problems with my graphics card. When I play games, 5-10 minutes into the game heavy artifacting occurs and the system freezes seconds later.

This happened 3 times. The first two times everything was working fine after rebooting. The third time there was artifacting on the POST screen (random pixels and multicolored characters on the screen) and the windows boot screen and it BSOD'd before getting to the desktop. The crash was GRAPHICS_DRIVER_TDR_FAULT and was caused by nvlddmkm.sys. I left the computer off for a few minutes and tried starting it again and it is working fine.

The idle temperature is in the low 40s, same as it has always been. I looked at the load temperature yesterday and it was reasonable (60-70). The cooler is clean and the fan is working so I don't think it's overheating but I am afraid of damaging the card so I have not stressed it yet.

It shouldn't be a driver issue because I have had these drivers installed for at least a month and everything has been stable until today.

Is the card dying? Any suggestions?

Nvidia Geforce 9800 GTX+
AMD Athlon II X4 620
4 GB G.Skill RAM
Gigabyte GA-MA770T-UD3P Motherboard
4 answers Last reply Best Answer
More about artifacting crash
  1. Best answer
    While your GPU may not be overheating, it looks like the video RAM is, overheating video RAM is the usual cause for artifacting. Download GPU-Z and run that in the background, it monitors temps for the various parts of the video card, including the video RAM. If the video RAM is overheating, you may have to remove the card's cooler, reapply the thermal paste to all the appropriate areas, then reattach the cooler. Some of the older nvidia cards out there are having problems with the thermal paste losing its effectiveness over time, leading to overheating issues, even if the cooler is perfectly clean.

    If there are no heat issues, it does seem that your card may be on its way out.
  2. GPU-Z is not showing RAM temperature. I do not think this card has temperature sensors on the RAM chips. In any case, the heatsink makes contact with the RAM chips using thermal pads. It won't touch the chips if I take off the pads because they are quite thick and there will be a gap between the heatsink and the chips. Is there another way I can confirm that the VRAM is the problem? Where can I get replacement thermal pads?
  3. It seems like the VRAM was overheating. I replaced the stock cooler with an Accelero L2 I had laying around and bought a pack of VRAM and MOSFET heatsinks. Everything seems to be fine now. I guess the thermal pads on the stock cooler degraded with age and became ineffective.
  4. Best answer selected by raptorjesus_76.
Ask a new question

Read More

Nvidia Graphics