Graphics card dying

[Didge3]

My trusty nVidia 8800 GTS seems to have kicked the bucket tonight

It all started a few days ago when the whole screen on my computer would freeze for about 10 - 30 seconds and then jump back into life exclaiming that the display driver stopped responding and has recovered, after this occurring several times I checked for new drivers and decided to roll back to a previous driver after a quick google search suggested. However anything graphically intensive (even Windows Aero effects) can make the card lock up. I'm thinking this is a hardware fault, the internet is littered with failed 8800's and it seems nVidia have even had law suits filed against them for ignoring a design flaw on this card.

Any suggestions other than to just buy a new card? I've manually turned the fan speed up (not that I feel it's overheated at 55C) I might underclock the card and see if that gets me anywhere?

Thanks

 

[Roadhoguk]

Usually when a Graphics card starts to die you will experience "Sharding" or tears" across the games or programs that you are running, where are you getting the temp reading from, if its freezing it may be down to undervoltage, have you checked the 12v Rail from the PSU and is it delivering the required Volts, check either with a multimeter or the BIOS screen, the best way to check is to put the card into another machine and try it there but if it is running at only 55c then it may well be another problem causing this.

 

[Didge3]

Usually when a Graphics card starts to die you will experience "Sharding" or tears" across the games or programs that you are running, where are you getting the temp reading from, if its freezing it may be down to undervoltage, have you checked the 12v Rail from the PSU and is it delivering the required Volts, check either with a multimeter or the BIOS screen, the best way to check is to put the card into another machine and try it there but if it is running at only 55c then it may well be another problem causing this.

Temp is from a range of sources, GPU-Z, CPUID HWMonitor and RivaTuner are the monitoring tools I use, my PSU is only about 5 months old and the card is getting 1.3V (according to all the monitoring software) which is what the card is rated at. I put my old graphics card in today and it seemed to work fine (although games run slower due to its lower performance) so I really think the card has a fault.
The screen doesn't often artifact but generally just goes black and after about 10 - 15 seconds my monitor light starts pulsing (no video signal), to recover from this I have to cut the power and reboot. Occasionally it will BSOD with an error along the lines of the video driver stopped responding and failed to restart (something like that).

I don't have any other PCI-E machines available except my Mac Pro but it won't work in that for obvious reasons.

 

[Roadhoguk]

video driver stopped responding and failed to restart

Yeah, that message effected quite a few cards from a couple of updates ago, although usually it restarts, yours sound a bit more serious if its BSOD`ing on you, do you get or catch sight of the error messages like "IRQ not less nor equal to" or "Kernel power 41" type messages ?.

 

[Didge3]

Yeah, that message effected quite a few cards from a couple of updates ago, although usually it restarts, yours sound a bit more serious if its BSOD`ing on you, do you get or catch sight of the error messages like "IRQ not less nor equal to" or "Kernel power 41" type messages ?.

I have automatic restart disabled when my computer BSOD's so I can read the errors, the error is what I said above about the video driver not responding.
IRQ not less nor equal and familiar errors like that I have only ever got when I've overclocked my processor too far. Next time it BSOD's I will make note of exactly what it says.