Video RAM artifacts

Hi, I've been recently getting some video ram artifacts in certain games, but not in other ones...was wondering why this was happening.

Ive had this computer for maybe 3 years now, I've only started seeing this towards the end of last year; in a couple games, I see weird lines coming usually off of guns in CS source, and in Splinter Cell the original, sometimes I get just big patches of red, blue, and green geometrical shapes. In TrackMania nations forever Ive also gotten weird artifacts.

I've also had weird lines in Far Cry 2, so its not just older games.

Yet certain other games run without any of these artifacts.

I've done a lot of searching around, and it seems the main causes of this are either bad Video ram, or overheating cards. My card seems to run from between 60 and 80 C; 60 when no games, and 70-80 when i have been gaming in source for 10-20 minutes.

So i am not entirely sure how I can fix this problem, hopefully I will not need a new video card. Its very odd how it only happens in some games, not others, and it never happens outside of the games. Like, I never have artifacts in the desktop environment, during web surfing or anything, except for certain games, and usually after they've been running a while. So not like initially usually.

Any ideas?

By the way, my graphics card is the NVIDIA Geforce 7900 GS. Ive already tried installing the latest driver and rebooting, no luck. However, I didnt uninstall the earlier one yet.

 

Some games, new or old, will work the video card (and its RAM) in different ways, or will cause it to overheat quicker. First, I'd pull the card (unplug the power cord from the PC first) and give it a good cleaning: the fan and heatsink especially, and make sure the fan is spinning free and easy. It should have a little 'bounce back': if you flick it lightly with your finger, it should kind of 'bounce back' a little bit when it stops spinning. If you feel comfortable with it, remove the heat sink and clean it, and clean the GPU, and apply a fresh layer of quality thermal snot. Then replace the heat sink firmly or find a decent 3rd party replacement for the cooler. Also, check your power supply and be sure it's kicking out stable voltages. HWmonitor will display all your voltages and system temps and can be a good guide to what's going on in your PC. It may be a good idea to run GPUZ in the background while gaming (click the Sensors tab and check the two little boxes at the bottom; a log will be saved to My Documents). This will record your video temps. If the card is idling at 60c, that is WAY too hot. I have an NVidia GTX260 and it never gets over 55c during crazy gaming sessions, and it idles at 40-42c.

 

I am a bit hesitant to do this, do you have more detailed instructions somewhere on how exactly to clean this kind of hardware? I have rarely ever opened my CPU case, and I wouldn't exactly know what to do, wouldn't want to risk damaging anything. I am somewhat of a hardware newbie you might say, when it comes to these things.

As for the software you recommended, I downloaded both programs and ran them. HWMonitor showed my Idle temp to be at 55 C, and I couldn't find info on the voltage with that tool for some reason.

For the GPU-Z, i had it save a log, while I did like 10-20 minutes of CS Source gaming. Turns out everything froze up and I got a BSOD at that point, DRIVER_NOT_LESS_OR_EQUAL pointing to cmdhelp.sys, which is related to my Comodo Internet Security program (I disabled both the firewall and Defense+ parts of it, so im not sure why it would crash like that).

First time I have gotten a BSOD with Comodo that I am aware of.

Well, according to the GPU-Z log, CPU Core Clock was a constant of 450 MHZ throughout, the GPU Memory clock was 660 MHZ, the Temp ranged from 55, before I ran the game, peaking at 74 C at one point.

The fan speed alternated between 20 and 40 %.

Thanks for the help so far, I appreciate it a lot.

 

If you have older hardware, HWMonitor may not be recognizing the motherboard sensors. Not a big problem. As to the BSOD: I'd head over to the Comodo Forums and see if there's an answer there. Seeing the fan speeds so low is probably the root of the issue. You can use RivaTuner to set the fan speed at higher rate; try 50% or 60%. I have my NVidia GTX260 set at 70% and it never gets over 60c during crazy gaming sessions (usually around 55-57c), and it idles at 40-43c. If you need help with RivaTuner, just Google it. RivaTuner is VERY popular and well known, and how-to guides can be found all over the place. For cleaning the video card: if you're not comfortable with removing the heatsink and fan, that's OK, but you should at least make sure the fan and heatsink are clean from dust and built-up crud. Canned air works well, and a QTip is good for removing stubborn caked-on dust. Check the fan as I described above and make sure it's spinning OK. Check and clean the CPU fan and heatsink also while you're in their. You can do all this without having to take anything apart other than the side off the case, and the power cord from the back of the PC.

Good luck! Let us know how it goes!

[dlb]

 

I'm just worried I will mess something up, I've never opened my case before really. I dont even know where the heatsink is

Im wondering if my graphics card itself is just bad and needs replacing. Im not sure if theres anyway to tell this.

I guess if I tried to clean the dust out, and cooled it better, and it still glitched, then it would be the card?

Is there a tool to test for bad Video ram?