Nvlddmkm, please help

[robersdee]

Hello TH members


after a week of troubleshooting I'm at the end of my rope

Some may recall an issue I had at the start of last week, I fired up Sniper elite Nazi Zombies, in the rendered cutscene at the beggining the FPS was atrociuos, to he point where it eventually froze (just when hitler looks into the camera, an image I shall remember for a long time after the week Ive had)

The screen then powered off and back on and I was greeted by this error

Display driver nvlddmkm stopped responding and has successfully recovered.

since that happened, this error happens randomly every 10-30 seconds while on the desktop

Theres hundreds of hreads (including a 151 page thread on the nvidia forums) about this, Ive tried every fix I can think of.

I'm led to believe this could be hardware, I'm just stuck as to which part, maybe you guys can help


here is a screenshot of gpuz afterburner and sysfan, the spike in the gpu usage is what happens, this will continue untill the mouse is moved then the screen will power off and I recieve the "Display driver nvlddmkm stopped responding and has successfully recovered." notification int he taskbar

things I have tried over many many many hours this week

* Reinstalling windows 3 times (each time the system can last upto 2 to 3 hours before randomly getting the error again, then it will do it constantly

* manually updating the nvlddmkm in system32/drivers

* clean installs of every driver revision from the last 3 months

* setting power to max performance (basically every fix detailed in the first 3 pages of google results)

* switching out the 680 for my old card ( same issue kept happening)

* reseating all cables and cards several times

* changing from my pcie power connecters for the supplied molex adapters


and probably more, Ive spent countless hours this week trying everything I can think of

So process of elimination it could be the 680 the psu or the mobo, surely it cant be the card as the problem persists when swapping out? and wouldnt I get more errors if it was the mobo?

so the psu? any sure fire way to test this? I dont have a physical tester, any what would be faulty? surely there would be other symptoms

if anyone has any ******* clue I've really lost the plot trying to fix this


cheers
Rob

 

[Gundy]

Wow. Sounds like a bad week. It's possible, however unlikely, that it could be a hard drive issue. If the driver files are being corrupted (or any of the .dll's or system resources that the driver relies on). But I would think that you would see issue with other programs as well if this were the case. If you have an old HDD lying around to test it out it might be worth it just to rule this failure mode out.


The only way to really test the PSU would be to hook an oscilliscope up to it with a current probe. Unless you have that already, don't bother. Replacing the PSU and seeing if the problem happens again is really the only way to be sure. One more thing: does it matter if you leave it overnight? Could be some sort of thermal issue in the PSU that's messing with one of the rails.

 

[robersdee]

bizarrely this is solved and may help others

it was malware, a bitcoin farming malware was using my card when it thought I wasnt there (absense of mouse movement)

so unlucky how this was downloaded 3 times on 3 different installations of windows. solved now anyway

this could help many others as this is a big issue and the particular malware (iehighutil.exe) is not flagged by most virus scans at the moment