nvlddmkm.sys and/or dxgkrnl.exe and/or broken card

[AmanoTC]

Hello guys (and girls?)! I am new to the forums as a registered user but I've been reading the forums on certain occasions.
First of all, english is not my mother tongue so please bear with me if I am writing something wrong.

I am posting here because I am looking for help or someone to point me in the right direction.

I was playing an old game on sunday morning (Heroes of Might & Magic 5) when suddenly I had artifacts all over my screen and I got the infamous nvlddmkm.sys-BSoD with Stop error 116 and with looping sound.

This was the first BSoD on my Windows 7 Home 64bit since approx. 4 months. Last fresh installation was in October 2009.

Since that BSoD i can reproduce this crash with a guarantee of 100% everytime I start a 3D-application / a game.
Desktop / 2D works fine, no problems at all.

So I started testing yesterday.

When I am alt-tabbing to desktop as soon as the artifacts / graphical glitches appear, I can see that I am having a TDR error: nvlddmkm.sys stopped working and has been restarted.
When I am not alt-tabbing fast enough, I get the BSoD.

So I took a look at my dmp-file with NirSoft Bluescreen View and saw that I not only got an error with nvlddmkm.sys but also with dxgkrnl.exe.

Next step was to try and update my DirectX but I already had the newest version and no files were replaced, thus I think that no files are corrupt.

I updated my NVIDIA drivers from 197.13 WHQL to 197.45 WHQL but that did not help at all.
I uninstalled the drivers, used Driver Sweeper from 3D Guru to completely remove any NVIDIA files that were not uninstalled (although I didn't do it in safe but in normal mode) and reinstalled 197.45. Error still persists.

I ran Memtest86+ v4.10 for 9 hours (5 passes) last night, no error, so I am pretty sure that my RAM is okay.

Neither RAM nor graphic card are overclocked. My timings for the RAM is 5-5-5-18 and 1.85V. Corsair says I shall use 5-5-5-18 and 1.8V but I can not set a lower number for Volt than 1.85 in my BIOS. Nevertheless I don't think that the difference between 1.85v and 1.8v is the problem since my system was stable and running 24/7 since October 2009.

My PSU is a beQuiet 750W so this is definately enough power to run the card.

I watched my temperatures, voltage, mhz etc. with Everest, Riva Tuner and GPU-Z. voltage and mhz were all as recommended by the manufacturer / Nvidia and idle temperature was 61°C (142°F), rising to 68°C (154.4°F) as soon as I started a game. My graphic card was approx. 10°C (50°F) cooler in January when I last cleaned my PC, so I guess the fan (and everything else) needs to be cleaned again, but nevertheless the temps are still low enough that overheating seems not to be the problem. CPU is at 30°C (86°F) while idle, btw.

A few minutes ago I used FurMark with the same result: as soon as the "fur-ring" appeared -> TDR error.

I have to add that I already had a 8800GTX and that card broke 3 months after the 2-year-warranty ended. At that time I bought a new card - a 8800GTS - which is running in my rig now.
As the old GTX broke, it was a broken VRAM and the optical symptoms were the same! Artifacts in 3D, TDR errors etc BUT after a few days I also had those random artifacts in 2D, even in BIOS. So I got the card checked and indeed it was a broken VRAM.
So this is the reason why I am here. With the described problem and the testing I already done I am leaning to another broken VRAM, another broken card as the reason of my problem. Which would suck because - can you guess? - warranty on the card ended in April
I was planning to buy a complete new system at the end of the year so to buy a new video card was not was I wanted to do.

Of course I hope that maybe the problem is not a hardware problem but a software problem and this is what I am looking for in your answers. Software problem oder broken card? Because this problem never appears while in 2D-/Desktop-mode but everytime I switch to 3D-/Gaming-mode.

I you need any more info (or screenshots or attached dmp-files) please let me know.
I really appreciate your time reading this and your answers! Thank you in advance.

 

[Tews]

There are several things that you can try..

1. run sfc /scannow from an elevated command prompt to verify the integrity of your system files..

2. Do a system restore to a point in time before you started having problems..

3. See if you can borrow a different gfx card just to test.... though a bad card is likely..

4. Do a repair install...

Let us know if you need further help and keep us informed..

 

[AmanoTC]

Hello Tews, thank you for your answer.

to 1.: Windows Resource Protection did not find any integrity violations.
So no errors.

to 2.: I wanted to do a system restore yesterday but the restore failed with the message "...a file was in use / locked, probably due to antivirus software...". So I disabled my Norton Antivirus, repeated the system restore but with the same error message. Any hint how I can completely block Norton to interfere with my system restore without deinstalling it?

to 3.: borrowing a card will be difficult.. But I will try to get my card installed in another computer. If I get the same error on the other machine, it is the card.

to 4.: will try and keep you updated.

 

[AmanoTC]

Update:
Neither repair install nor system restore helped.
Quite the contrary, the crashes now even happen while playing a video in fullscreen mode and just a few minutes ago while browsing (with Aero active)
Card gets checked on friday.

 

[AmanoTC]

Hey guys!
Sorry, I totally forgot about you.

To let you know, all my assumptions were correct. The card - better said, the VRAM - was broken.
So I had to get me a new card. Since I will buy a complete new and uptodate rig at the end of the year, I got myself a Radeon 5670 to replace my 8800GTS.

Thank you for all the help you gave me. I appreciate it!