Graphic corruption... what is causing this??

[Hicks12]

Hi guys, wondering if you could help me out as I have experienced this numerous times and it seems rather random in occurrence, a lot of the times its videos that causes it but I really cant pin point the cause.



Really annoying as a reboot seems to fix it no problem and having dual monitors it seems limited to one monitor at a time but its completely random as to what monitor experiences the problem!

As of right now I have the issue on my "main" monitor, if I unplug it the issue switches to my other monitor, main is connected via Display port and hte other dvi.

Any ideas? Really annoying issue!

[Jay]

looks like it could be video memory overheating?

[eyezpy]

It could also mean your video card is dying...
I would give everything in the the case a good clean to make sure no dust is causing over heating by blocking the fan or heatsink, make sure everything is plugged in securely and that all the fan is working. Then do a stress test with something like furmark for a few hours and see if the problem comes up.
If it does than its a good chance its the GPU and to be sure I would do the same test with the video card in another PC that you know has no problems.

[Bonebreaker777]

Eyezpy got a point, do tests, all which are available. That card is on it's way out most likely.
If you have warranty, try to initiate RMA procedures, if not, in similar cases re-baking the card helped me (but try that as last resort of course).

[kalniel]

Your GTX9800 had the infamous G92 chip which suffered from solder issues, very unlikely other chips will have the same thing.

[kalniel]

Would this kind of thing still work for new GPUs though? I mean, you never really know where the issue is right?

Unlikely - you might be doing more damage than you cure. The G92 bump issue was fairly quickly diagnosed because it caused failures in laptops whose chips go through several heat cycles.

[Bonebreaker777]

Recently a out of warranty MSI 460 Lightning was cured in the same way, so I wouldn't dismiss this option straight away.

[Kiyam]

Same method worked when a motherboard I had was having RAM socket problems, seems like a little bit of heat goes a long way :3

[kalniel]

I stand corrected then if hexites have personal experience of this working with other cards

[spiritdreams]

Common thing in fixing computer hardware, it is actually a process called reflow, used when solder joints has become loose on boards.

Depending on the solder used, 150C - 170C reflow temperature. usually 5-10mins

Word of warning, plastic connectors points will melt in most cases, wrap them in foil helps, also capacitor degrades very quickly at the temperature, so check for bloated capacitors after the "bake".

Finally all stickers will come off during the reflow process, if you have not removed them yet, this will void any warranty you may have had.

[koyejo]

could be 1 of many things
1. psu not supplying enough power
2. gpu overheating
3. gpu dying

[Thorsson]

Hmm as it's only one monitor of two then it doesn't sound like the GPU. Best would be to test with one monitor and a different cable and see if it still happens. First though I'd take the video card out and check it for dust.

[Bonebreaker777]

Maybe is THE specific output?
More tests required! Test whatever you can.
Same cables but with reductions and different outputs?

[Honosuseri]

First off have you tried returning to stock speeds? 900Mhz doesn't sound like a factory OC to me fella, did you bump the VRAM too? I did notice that running dual screens bumped the idle temps up on my EVGA 1GB 460.

Also the new Nvidia drivers caused no end of hassle for me with weird issues and the rig crashing without a BSOD. I'm currently on 314.22 which is stable. I really hope it is as simple as that to solve your issues

[Bonebreaker777]

Change drivers? And ye, return to default clock, please.

Semi damaged RAMDAC?