Display driver nvlddmkm stopped responding...

[KevSC1]

This problem started for me a few days ago out of the blue. I hadn't recently installed any new programs or drivers. It began with just the screen freezing for a second and then returning to normal with the error message "Display driver nvlddmkm stopped responding and has successfully recovered." I was running an EVGA NVIDIA 8800GTS 640MB card at the time. Nothing on the computer is overclocked nor has it been. Then the next day the error started to happen more frequently so I upgraded the drivers to the most recent release 191.07 WHQL. After this I started to see a lot of screen artifacts and began to run some diagnostic tools like FurMark, each time I tried to run this it would eventually freeze up the OS, resulting in either a BSOD with nvlddmkm.sys (0x00000116) to blame or just freeze up the screen and never come back until a system reboot. All of the temperatures were always within operating ranges. I tried cleaning out the drivers and installing various releases, as well as pretty much every posted solution I could find, but nothing helped. Attached are some of the pictures of the artifacts I was seeing. I decided to contact EVGA who told me to RMA the card as it was under warranty. Since there were graphic problems in the BIOS they thought it had to be the card malfunctioning and not the drivers, etc.

I thought all would be well and good from here, So i popped in the only spare graphics card I have which is a BFG NVIDIA 6200OC PCI. Unfortunately the problem continued to some degree. I no longer have the graphic artifacts, but I am still getting the "Display driver nvlddmkm stopped responding and has successfully recovered." error whenever I try to do anything even minimally graphics intensive, for example looking though the pictures I posted at a fast pace, or even run the Windows Vista performance rating diagnostic. I decided to go ahead and do a fresh install of Vista, and even with that I am still getting the error message. Note that the OS was current with all updates in both cases.

I was hoping it was simply a defective graphics card, but even with a new card and a clean install of the OS, I am still getting the error message. All of the images are from when the 8800GTS was installed. I have done all the basic stuff like checking the HDD, RAM, virus scan, etc. Does anyone have any thoughts as to what could be causing this problem? Please let me know if any other information might be helpful.

Thanks,
KevSC1

DXDiag can be viewed here: Share Text simply - ShareText.org

Images of the artifacts:


















System Specs:

Motherboard: Asus Striker Extreme
Processor: Intel Core 2 Quad Core Q6600 @2.4GHz
Memory: 4GB Patriot (2 x 2GB) DDR2 800, PDC24G6400ELK
Hard Drive: 2x Samsung 500GB HD501LJ, and 2x Samsung 500GB HD501LJ in Raid 1
Graphics Card: BFG NVIDIA 6200OC 256MB PCI (Was EVGA NVIDIA 8800GTS 640MB PCIe)
OS: Windows Vista 32-Bit SP2
Sound Card: Onboard, SoundMax
Power Supply: Thermal Take 700W W0106RU
Cooling: Swiftech H20-120 CPU Cooling Kit With Apogee GT
Monitor: Acer 22" Widewcreen X223W, Running at 1680x1050
Browser: FF/IE8
Optical: 2x LITE-ON DVD-RW SATA LH-20A1L

 

[DocHoliday19]

I am having a VERY VERY similar problem myself. A few days ago while playing a game i got an error saying n4v_disp has stopped working and my graphics became all pixilated like in your screenshots. I updated and reinstalled drivers, switched card slots on my mobo, tried everything possible related to the videocard but nothing worked. I am currently RMA'ing the card (9800 GT) and am using my spare 8600 GTS. For the first few days of using the replacement card things were fine now all of a sudden my games will freeze but i can pull up task manager and end them. Once i do my screen justs freezes and i have to reboot and i am now having random pixilation like in your screenshots. Not nearly as bad as my other card though. I couldnt even boot up anymore with the 9800.

The only things we have that are exactly the same are the Processor and Monitor.

There is a new beta driver out as of a few days ago, i will be trying this - NVIDIA DRIVERS 195.55

Hopefully we can help eachother and get some outside help!

 

[Slaymate]

I've had this same problem numerous times in the past years. It can be caused by a lot of different things. When your playing a game the video card is the hardest working component in the pc. So if some component starts to produce the occasional error 2 things will usually happen. 1) the game will crash or freeze 2) the video driver will crash or freeze.

Start by cleaning out your case with compressed air. If need be remove and clean each component individually with compressed air. If you remove the heatsinks from the cpu and gpu then it is alright to use alcohol (non-drinking type) to clean off the old paste but be sure to use a good quality paste when re-assembling the heatsink.

Once you have everything clean and we know dust isn't causing any heat issues then it's time to start monitoring your hardware to see if old age or if your like me, abuse has taken a toll. HWMonitor will show you a lot of stuff about your hardware's health. Leave it running when your testing and playing as it will keep track of the temps, volts, fan speeds and such of various components with a High, Low and Realtime (at this very moment) of each.
HWMonitor

Use a program like Furmark to stress test your video card. Install it and run the stress test for a couple of minutes and report the starting Temp and the Temp when you stop the test.
FurMark v1.6.5 download from Guru3D.com

You can use a program like OCCT to stress test your cpu. It will run for 1 hour by default unless it detects an error. Once again report back with your Temps and or any errors reported.
Download - OCCT Website english

Use Memtest to test your memory. The download is a .iso file that you'll have to burn to a CD. And it doesn't run in Windows so you can't record any temps. Let it run thru all of it's tests 2 or 3 times and report back if it produces errors.
Memtest86+ - Advanced Memory Diagnostic Tool

 

[KevSC1]

Thanks for the reply,

Start by cleaning out your case with compressed air. If need be remove and clean each component individually with compressed air. If you remove the heatsinks from the cpu and gpu then it is alright to use alcohol (non-drinking type) to clean off the old paste but be sure to use a good quality paste when re-assembling the heatsink.

I keep my case pretty clean, giving it a good dusting and checking the fans every few weeks or so, so this should not be a problem. When the comp was put together I used good quality thermal paste, one concern might be the NB and SB, as I never messed with the stock paste that ASUS had applied, which I hear is not that great.

Once you have everything clean and we know dust isn't causing any heat issues then it's time to start monitoring your hardware to see if old age or if your like me, abuse has taken a toll. HWMonitor will show you a lot of stuff about your hardware's health. Leave it running when your testing and playing as it will keep track of the temps, volts, fan speeds and such of various components with a High, Low and Realtime (at this very moment) of each.
HWMonitor

I've been using Everest, the temps all seem within range, the highest ones that are of possible concern are the NB at 52 generally and the SB at 66 at load.

Use a program like Furmark to stress test your video card. Install it and run the stress test for a couple of minutes and report the starting Temp and the Temp when you stop the test.
FurMark v1.6.5 download from Guru3D.com

I used FurMark with my 8800GTS and would get perhaps a minute before the artifacts/BSOD, now with this card I can't even get more than a few seconds before the driver crashes and the screen goes all scrambled.

Current example of what happens:

This is from running FurMark or something similarly graphics intensive, here the computer completely freezes and has to be rebooted manually. If I even dare ran a game it would just crash the driver and recover with the game still running every 15 seconds. If I am using the windows performance diagnostic to rate the computer then it will show this kind of screen for a few seconds and then recover back to the desktop showing the error message "Display driver nvlddmkm stopped responding and has successfully recovered."

You can use a program like OCCT to stress test your cpu. It will run for 1 hour by default unless it detects an error. Once again report back with your Temps and or any errors reported.
Download - OCCT Website english

I have checked the CPU wi th various burn in tests including Prime95, IntelBurnTest, Passmark Burn In Test, OCCT, and such and all have come up fine and without errors.

Use Memtest to test your memory. The download is a .iso file that you'll have to burn to a CD. And it doesn't run in Windows so you can't record any temps. Let it run thru all of it's tests 2 or 3 times and report back if it produces errors.
Memtest86+ - Advanced Memory Diagnostic Tool

I have run several memory diagnostic tools overnight including Memtest, and all results were fine, I also tried using only one of each stick of ram and different slots, and no luck there.


Thanks,
Kevan

 

[KevSC1]

Update:

I decided to go ahead and try installing Windows XP on another HD to try and see if I would get any errors there. I installed XP and got all the Microsoft updates and latest drives and then ran FurMark. It ran without any problems at 640x480, 1680x1050, etc and even with 16x AA on. In my Vista installation the computer will freeze up after about 20 seconds at 640x480 without AA. I did notice one type of graphical problem in the Windows XP installation however, the images are attached below, these were simply screen shots taken from within Windows.

Screen Shots taken in Windows XP via Print Screen:













Does anyone have any further thoughts as to why I am having trouble in Windows Vista but not XP? I am pretty sure I have tried all the conventional fixes for the nvlddmkm.sys error. The motherboard is still under warranty and it has been suggested that this might be the root of the problem, but since I was not able to replicate the problem in XP, but getting artifacts or BSODs I am weary when it comes to RMA'ing the motherboard without any hard proof that it is the cause.

Thanks,
Kevan

 

[Slaymate]

If Furmark is crashing your PC then your video cards are bad. You can try taking them apart and cleaning them, it might help.

 

[KevSC1]

If Furmark is crashing your PC then your video cards are bad. You can try taking them apart and cleaning them, it might help.

Two bad video cards seems unlikely since this 6200OC is brand new, could the mobo perhaps be damaging the cards in some way?

Kevan

 

[solidsteve]

if its any help whats so ever (which it most likely wont be for now) my gtx 295 is giveing me the same sort of error

i have had the error that the graphics display driver stoped working and have been playing games with it and they have frozen up specially in mirrors edge with the physx.

i'm just about to give my pc a clean and stress test ect and see what hapens, have been keeping a eye on the temps and they seem fine.

by the way i installed folding@home could that have anything to do with it.

 

[Slaymate]

I can run my video cards on Furmark without them crashing or artifacting. Brand new cards are defective all the time. Try using a tool like Rivatuner or EVGA precision to "reduce" the clock speed of your card and see if it quit's artifacting or failing Furmark.

The video card plugs into the mobo so it is possible that the mobo could be bad, but not likely. It could be a bad connection between the card and the mobo. Make sure the weight of the card isn't flexing the card. Try holding the back edge of the card up and see if that helps.

 

[KevSC1]

So the problem has been fixed...for now. I decided to go ahead an see if the 6200OC could be faulty as well so I picked up an NVIDIA GeForce 210 512MB while I await my RMA replacement for the 8800GTS (a GT 250 512MB). Indeed my luck is just horrible, the 8800GTS died and then the 6200OC just happened to have been a bad card from the start, now all is well with the 210.

Thanks,
Kevan