• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

The dreaded Display driver nvlddmkm stopped responding and has successfully recovered

mikeo

mikeo

mikeo

Soldato
Joined
17 Aug 2005
Posts
2,828
Location
SW Scotland
Now I know there are a number of threads on here about this issue and on other forums (Nvidia forum has a 50+ page thread!!!).

Since building a new PC round Vista premium (64 bit) I've had this happen a number of times. Screen goes black (sound carries on if it's a game) and eventually drops to the desktop saying the usual "recovered from display driver blah blah blah). Though a couple of times, it's happen just in normal windows tasks (EG. while browsing etc.).

Now been building PC's off and on for some 20+ years (yes I am that old) so have a fair idea of what I'm doing ... but this is a very irritating problem.

Besides the CPU/mobo/memory, all other components have been running fine in an XP based machine previously for a year + (and NO I don't want to go back to XP, as I rather like Vista). All drivers up-to-date, OS patched to latest state. All usual stress tests run OK (Memtest+, ORTHOS, looping 3Dmark Vantage) and pass fine. CPU/GPU temps also fine, both at rest and under load (checked via RealTemp and RivaTuner). No overclocked components (yet) other than factory OC'd BFG 8800 GTS and that's been fine for over a year in the previous PC. PSU is more than up to the job, 800w Tagan modular.

I've seen loads of things blamed... memory, Vista, Graphics card, AERO, Overclocking etc. etc. But in my case, can't seem to pin it down.

One "interesting" post on a forum, was someone saying that his graphics card vendor had suggested he pulled the card and checked the screws holding the HSF were all tight, even if temps "appeared" fine. Out of interest I pulled mine and checked and yes, they were not as tight as they could be. So tightened them up... not like I think this is the problem. Only time will tell, as the problem appears random.

Have noticed though, if it happens, then even if it appears to recover, the problem will happen straight away if going back into a game. But if you switch off, give it a second or two and then switch back on, it then carries on OK (for a while). Almost like something has not been "reset" unless you close the PC right down. This sounds more like a software problem to me, rather than hardware.

Also seen someone else with the theory that it is because of multiple copies of nvlddmkm on your machine (which there are). Though he seemed to think that it was because of several different updates of the graphics card drivers (IE. old stuff not being uninstalled correctly). Well I can't see it is, as my machine appears to have several different copies of nvlddmkm hanging around and I've only installed the latest driver "once" when I built the machine.


Sorry for rambling on, but this is getting up my t*ts! and it's also to stop loads of posts suggesting checking temps etc. (as well meaning as they are).

So...

- Anyone got any current info / ideas on this?

- Is it more associated with earlier Nvidia cards (8800 and before)? As I would upgrade to a 200 series card if I thought it would get rid of this!

- Are ATI cards effected less? (obviously a different driver here but TDR still applies). As again I would be willing to move to a 4850/70 to sort it.

Thanks folks.
 
Last edited:
I also get this really annoying issue myself, and system works fine from different tests etc. that can be run to find errors, with no apparent overheating issues or fluctuating spikes.

In particular I get the issue with call of duty 4.

Hopefully someone will come up with a helpful suggestion because it is a real pain and incredibly random when it occurs. It is incredibly frustrating, so much so, that I myself am considering a move to ATI as I never experienced such issues with them in the past, even though I prefer the performance of the nvidia cards. I do feel as though its an nvidia driver issue, sadly my random googling in the past has not been able to find a fix to the issue :(
 
ive not had this issue with the new nivda drivers 177.83 i thinks using vista home prem 64 use driver cleaner to remove your old gfx drivers and isntall in safe mode hope this helps
 
scooby555t

Thanks... but the latest WHQL for 8800 series cards under Vista 64, are 175.19 (which I'm running). And do not the later drivers just add support for the later series cards (IE. 9800, 200 series)?
 
Having just read the posts on guru3d.com about the 177.83, I think I'll give them a miss at the moment.
 
I've had this problem occur a grand total of once since I built my PC last year (hardware has stayed the same ever since). It was a few weeks ago while playing TF2, and it occured whilst using the 175.19 drivers - the set that you're currently using. I'm using 64-bit Vista aswell, with a 8800GTS.

I've since updated to 177.79 (the driver released shortly before the Physx driver package), and I've not had it reoccur. Like I said though, only ever happened once... can only assume the driver was to blame. So, I guess I'd recommend trying a newer set (such as the aforementioned 177.79):

http://www.nvidia.co.uk/object/winvista_x64_177.79_beta_uk.html
 
Last edited:
MiNxYyY

Not just related to Nvidia then I suspect. But then again, I was starting to think this anyway, as there's a BIG thread here about 48** cards and Vista 64.

hohum

Thanks for that, I see these appear to be classed as WHQL certified, though the actual download seems to mention beta. I'm downloading now. Should I continue to have problems, then I think I'll invest in the subscription version of DriverCleaner (rather than the rather dated freeware version still doing the rounds) and give them a try.

Interesting to see from much reading on the Internet, that this problem appears to be truly random (though obviously there are many reasons that can genuinely cause this problem). From folks like yourself, who only have it happen rarely. To folks like me, that have it occur a bit more frequently, To folks that have the problem so bad, as to make their PC's virtually unusable, especially for gaming. Also spotted a couple of folks that have dual boot systems (XP/Vista) and only get it on Vista. So in these cases, this would pretty much appear to rule out hardware problems (though I suppose not compatibility problems between Vista/hardware). No easy solution here by the looks.

I have also seen comments that it seems to effect a large number of people running 8800 series cards. But in my opinion, this could just be because these cards have now been around a while and hence have a pretty large user base.

Thanks folks for your contributions up to now.

Would be VERY interested to hear from anyone that had this problem and...

1) Moved to a later card, Nvidia or ATI and this appeared to sort it.

2) Or did something specific that they can be fairly sure sorted their problem.

Thanks again.
 
Last edited:
maybe try underclocking the graphics card, factory overclocked cards aren't always thoroughly tested I remember buying an XFX XXX and having nothing but problems with it.
 
Only times I've ever had this error is when one of the following in my system is overclocked too far - northbridge, ram, cpu - in that order... once they were stabalised never had the problem again. The NB has been the main cause in my case - making sure theres a fan on it so it doesn't get too warm and that the voltage is neither too high or too low - both extremes seem to cause it - worked wonders. From my experience the error is rarely related to the GPU itself unless the GPU is overheating and thats easy enough to check.
 
Lol i started that threat you mentioned about the same type of thing happening with 48xx cards.

It's extremely weird.

With all these Nvidia cards i've not had it happen on Vista 64-Bit with dual displays : 8500GT, 8800GTX, 9800GX2, GTX 260, and GTX 280.

But with ATI i got the problem, with both 4870 and 4870X2 (way way less on the X2 though, so rare it's not really a problem for me anymore).

If i was you i'd try a few different NV drivers, no harm in trying. I've tried about 20+ versions on Vista with all the NV cards i've had and not had it happen, so i'd use driver cleaner to make sure it's not something else.

From building about 100 cheap Vista computers to sell over the past year, i've found that with this problem on Vista, when it's not the drivers - it's nearly always faulty RAM. And XP will sometimes be stable with the same RAM, but Vista just stresses things more to bring out problems with faulty RAM.
I've even found the same with overclocking, i've highly overclocked computers before and they would be fine in XP, but in Vista i'd have to clock them down a bit, sometimes i'd get the most bizarre problems from this too (like Windows Update not working but everything else fine).
 
Last edited:
mmj uk

Good suggestion about under-clocking the graphics card. This is in fact on my list of things to try. My current card is a factory overclocked BFG 8800 GTS. Though it has run fine in XP for the best part of a year now. Though aware that I'm now obviously running a different OS. I'll await to see if it occurs again before trying anything else.

As I've tightened my screws (so to speak) on the HSF on the graphics card. I'll see if this has done anything first. Though RivaTuner seemed to imply that the card never goes much above 60C when gaming. The explanation I read, was that obviously monitoring software has a time interval between taking temp readings and that it could "in theory" spike higher in between sampling the temp. Not convinced myself as this has also happened in Windows itself, as opposed to in gaming. But as it's an easy option to try first and the screws holding the HSF were in fact a little slack.... Only time will tell, as the problem is pretty random (IE. could be 5 minutes, or 5 days).

Rroff

Overclocking in general does appear to be one of the main causes, from my now extensive reading (sad eh!). Nothing other than graphics card currently OC'd. Northbridge on these boards does appear to get a little warm, but mine does not appear to get "hot to the touch" as some have reported. This is one reason why I chose the Alpine 7 CPU cooler, as this blows down onto the CPU HeatSink and hence creates a fair old draft over the North Bridge (rather than across the HS as in the Freezer 7 pro). Also got CPU set in BIOS on 100% rather than automatic for the same reason. Suppose I could check the voltages though. I notice that you've got the same board as me, well the RAID version. Question... have you found NB voltages a problem area on this board?

Paradisco

Nope. Only one LG 1960TR monitor.

Mr.B

Thanks for your input. I'm only running a single monitor / card (IE. no SLI or x2 card), so looks like I'm just unlucky. Up to now it's only happened a handful of times, but enough that it rather spoils anything I'm doing, whether it be gaming, or something more series. As it's always at the back of your mind (that the problem may suddenly occur).

Again I agree that a lot of people point the finger at the RAM (quite rightly I suspect in a lot of cases). And I know only too well how a transient memory fault can drive you mad. In this case, I've done everything I can to eliminate the memory from the equation. I especially bought Corsair XMS2 800 memory with CL5 as this appeared to be one brand / type that no one reported problems with this mobo. And experience again tells me you are better picking memory that's not right up at the limit that the mobo can supposedly support. I've run both MEMTEST+ and ORTHOS for more hours than I can remember and no problems reported. Also realised today that Vista itself has its own memory diagnostic tool, so ran that for 5+ hours on the extended test and no problems reported. Now I know that even that does not 100% eliminate the memory, but it does push it to the back of the problem queue in my opinion.

If problem occurs again. Then I think my plan of action will be (basically in order of ease / cost):

- Try 177.83 drivers after running current up-to-date version of DriverCleaner.
- Disable AERO (something else I've seen blamed)
- Try under-clocking graphics card. Maybe set it back to standard speeds for the 8800 GTS might be worth a try.
- Consider buying a 260 GTX as they seem to have come down ludicrously in price (by then the + versions might even be around). A bit lost upgrade wise on my current 19in monitor, but it would then give me an excuse to buy a wide screen monitor!

This problem is a bit ironic really, as I've done nothing but champion Vista 64 to other folks. Saying things like "tut tut it's perfectly stable, there's nothing wrong with it, it's as stable as XP" . Whoops!!!

If anyone else wants to comment, then jump right in.

Thanks again.
 
Last edited:
i've had similar problems recently with my 8800 and vista 64, what was causing my problem was muh HD going into sleep mode, i'd hear one of my drives spin up just before the screen would go black for a few seconds and id get the error, then randomly id keep getting the error until i rebooted and it cleared. i just increased the hd sleep timer in the power settings and its stopped happening to me now.

Could just be i have a dodgy PSU though and the nvlddmkm error is a lame generic one and can be caused by many things
 
i used to get this error when vista first came out and i havent had it for ages
i put the error down to overclocking my CPU when i put it back to stock i havent had the error since
it does appear that every user with the error puts the problem down to something different.
ive read on many forums that the fix to this error is to delete the nvlddmkm file from the c:\windows\system32\drivers folder
i havent done this myself so i dont know if it works

also found this on a forum too:

Fix: Go to windows/system32/drivers and rename nvlddmkm.sys to
nvlddmkm.sys.old. Go to the nvidia directory and find the file nvlddmkm.sy_
and copy it to windows/system32. Using the cmd window (DOS box) type
EXPAND.EXE nvlddmkm.sy_ nvlddmkm.sys. When the expansion is complete, copy
the new nvlddmkm.sys to windows/system32/drivers and restart the computer.
 
Last edited:
Reznor

Know what you mean. I have HD set to never power down. I also don't like Intel Speedstep. So all that is disabled in the BIOS. Don't like the system, however clever it thinks it is, playing with multipliers / voltages etc. (that's just me).

Kmufc77

I had come across these before as well (though thanks for posting). Not sure I'm too keen to mess here. Though must admit that I'm not sure why there appear to be more than one copy of nvlddmkm.sys on my system (NB. have only installed one set of graphics card drivers when machine first built). They also have three "different" dates on them. Though to me, it looks like the latest version is in the right place.
 
I used to get this problem all the time, but found a fix that works, same principal as the one kmufc77 has mentioned above, replacing the nvlddmkm.sys file.

As far as i can work out, the problem is aggrevated by overclocking etc which is why this is usually blamed, but as far as i've seen (on a 7600, 3x8800gtx's and a 280) its down to NV's lazy coding.

when you uninstall the nvidia driver package, or install an updated driver over the top of an old one, the NV code doesnt remove everything as it should (this may have been fixed in the latest ones but i have no way to tell as i dont have the issue!)

to rectify the issue, uninstall the old driver then reboot in safe mode,
do a search for nvlddmkm.sys and delete (or rename as kmufc77 suggested) all 5-6 entries you'll find, (you may have fewer or a lot more, on my laptop i once had 14!!!)
anyway delete them all (or everything nv related if you want to be picky!)
and the unpacked install folder (by default c:\Nvidia) reboot again and install the new pakage.

this has worked for me every time and havent had this issue for the last 8 months since i've done it so cant complain!

hope this helps
 
You must log in or register to reply here.
Back
Top Bottom