Really weird (and bad) 8800 GTS problem

[crashed97tsi]

My Specs:
AMD 5600+
BioStar Tforce 590SLI
4GB Corsair XMS2 DDR2 800
Creative X-FI Extreme Music
EVGA Geforce 8800 GTS 320MB
3 Sata hard drives, some peripherals, and 4 cooling fans (120mm)
Coolermaster Realpower 550
Microsoft Windows Vista Ultimate

Earlier this morning, I was playing SH4 and everything was running fine. At some point the game did a hard lock up, which it does do in this game from time to time (Like once every four hours). I have to reset the machine, and instantly I notice something is wrong because the logo screen and the initial boot screen have rows of white pixels showing across the screen. The rows are more horizontal then vertical. However, once the boot process goes into my raid screen and Hardware monitor screen show upin the boot process, the pixels go away.

Then, when the vista load screen comes up (the one with the black background and the green loading bar), vertical rows of blue pixels show up. After that screen, the rest of the boot process has no pixels.

Once in Vista, the graphics card is unusable by the OS. It says Vista had to stop the device because it reported a problem (code 43). I searched around for some solutions or what this even means but so far I have come up empty. Does anyone have any tips, or perhaps a good diagnostic program I could run to see if anything is wrong with my card? Any help is greatly appreciated.

Thanks

 

[olesenster]

http://forums.nvidia.com/index.php?showtopic=28520&hl=

 

[Ibanezrg570]

Sounds like your card is artifacting. I would venture to say it was a driver problem if all you were recieving was that error in Vista, but, seeing as how you are seeing artifacts in boot-up screens before windows is loaded (and hense, before any of those graphics drivers have been loaded) it would seem as though the card is faulty.

If you are overclocking the card, try backing off the clock to see if it you can remedy the issue. If not, I would say RMA it as artifacts during boot up are almost guaranteed to be a bad card.

 

[crashed97tsi]

that's what I was afraid of. I'm putting my old card back in right now to confirm. This thing is only 4 months ol, too.

 

[crashed97tsi]

Alright, well I swapped out my 8800GTS for my 7950 GT. The 7950 eliminated the pixel problem. Man, I just must be having bad luck with components lately or QC is going out the window. 90% of the products I purchased in the last year have failed.....

 

[tamalero]

crashed: Then you need to take a lot at the rest of your components, or your electric stuff
you have a tri-fasic conector?
you have a regulator or a no-break?
cause lousy energy conditions can kill your system FAST

 

[nukemaster]

my vote goes for bad vram. since on P.O.S.T(power on self test) there are NO drivers loaded....when video ram goes you get artifacts like this....they slowly get worse....i had them(Geforce 4 Ti4200) in the bios(and on post) and not in windows....then slowly they made there way to windows....The rma'd card still runs great to this day....

Here is a link with some info on video card artifacts....
http://www.playtool.com/pages/artifacts/artifacts.html
i wish i kept my picture of it to show...but it was a long time ago....

RMA it....but first test it on another computer if you can....low power may cause this too....

 

[crashed97tsi]

I have a battery backup already because of the power (real old wiring in the building). I was thinking of upgrading to a better one soon, however this one has been working fine and it supposedly has a power conditioner type procedure to provide clean power (I do not know much about the subject, but it supposedly does that). The other things you mentioned I do not know what they are but I don't have them.

I always suspected power but everytime the maintenance guys comes in and checks the outlets everything is fine. That and most of the components that failed have been faulty from the getgo and purchased in my previous apartment (brand new wiring, worked flawlessly).

However, if you are willing to explain to me how I can check my AC lines to see if they are in spec(without killing myself ) I will be more than happy to listen

 

[nukemaster]

Your ups(battery backup) software may tell you the line voltage...

Or if you have a volt meter you can test that way....be careful though

 

[mitchellvii]

Check the latest BIOS for your MB. I had an out of date BIOS before and it did that.

 

[nel]

Or u can try to stick your finger in a outlet so that you get to feel how much voltage it has... lol

why not just get a conditioner or a surge protector... let just say your component are dying from the spike. then, why your power supply never get affected which is the only one that drawing a current from the wall... it sounds more like it OVERHEATED. Make sure all the heat in your case are all coming out and not just circulating inside...

 

[Neverknowsbest]

If you were getting hard locks while running a 3d application after hours of stable use with no real cause from the application you should have been thinking about a very serious problem right away.

Sounds like your card over heated and either fried a memory module or one of the pixel pipelines or cache memory on the main die. You might be able to recover if you force the memory and core clocks speeds lower. Although that usually only works when your slowly over clocking and waiting for artifacts to appear in stress test bench marks and then immediately drop the clock frequencies.