• The nvlddmkm error (big post, help me and others)
    30 replies, posted
Since there are so many of us with the 'Display driver nvlddmkm stopped responding and was recovered' (sometimes it pops up as 'Display Driver NVIDIA Kernal Mode Driver Version xxx.xx stopped responding and has successfully recovered' instead) issue (myself included), I think it would be worthwhile me posting everything I know about it from all the investigating I've been doing over the last few weeks. The best sources of help for this issue will probably be found in [url=http://forums.nvidia.com/index.php?showtopic=100800&hl=nvlddmkm]this[/url] thread on the nvidia forums and [url=http://nvidia.custhelp.com/cgi-bin/nvidia.cfg/php/enduser/std_adp.php?p_faqid=22]this[/url] post in the nvidia customer help pages. The first goes over many of the causes and fixes of the problem, the latter contains some motherboard BIOS settings which may help. There is also an entire website dedicated to this problem, but it hasn't been updated in a long time and contains tips with little more than anecdotal evidence for permanent fixes, but you could always have a look at it [url=http://www.nvlddmkm.com/]here[/url] anyway. I'm not going to waste time repeating what other sites recommend since that would take all day to type up, so I will post what I have tried with my system and I would advise other people with this problem to try the same things. Here are the specs of my PC since it'll probably help: i5 2500K 3.3GHz EVGA GTX 570 8GB Corsair XMS3 RAM MSI P67A-GD65 B3 mobo 1TB HDD Windows 7 Home Premium 64 bit Silverstone Strider 850W PSU Should probably add that this doesn't happen specifically at any time, sometimes games play totally fine and sometimes they don't play at all. One thing I can say is that it happens quite often when I try to watch videos. Things I have tried to no avail: My first GTX 570 ended up dying and stopping Windows from even getting past the loading screen and was RMA'd, EVGA sent me a new one. I tried an 8500GT and that worked ok but obviously it doesn't have the same power requirements as a 570. Checked RAM settings in BIOS and made sure they are set to the reccommended settings Totally uninstalled and removed driver remnants using Driver Sweeper and Driver Cleaner and reinstalled them Updated motherboard BIOS Checked temps, GPU barely goes above 70C when maxed out in crysis. The problem occurs even when my card idles around 35C. Apparently there can be driver conflicts between onboard sound card drivers and the nvidia sound driver, so I followed careful steps to remove the nvidia sound drivers which are explained [url=http://www.evga.com/forums/tm.aspx?m=586183]here[/url]. After that didn't help I even disabled my onboard audio in the BIOS, that made no difference. Ran memtest and got one pass and no errors Checked my card was seated properly. I was worried that it wasn't because some SATA ports are beneath the end of the card but there is definitely some clearance between the card and the ports. Somebody on the PC Gamer forums informed me that there is no temperature sensor on the memory chip and I underclocked the memory clock by 100MHz to see if the driver still crashed. Things I can try: [del]Leave memtest for a long time, see if I can get many passes with no errors[/del] ran it on each individual stick with 2 passes & no errors, then all 4 sticks with 4 passes & no errors [del]Fresh install of Windows? Not sure what that will achieve since Windows Update will install new drivers and updates for everything and there's a good chance I'll end up back where I am[/del] formatted my hdd, put windows on a ssd and all my stuff is now back on the hdd. still happening. Overclocking slightly in case it underclocks to an unstable level when idle Fully uninstalling the current drivers and installing the oldest compatible ones. Apparenly this has fixed the problem for some people. [del]Check that my hard drive isn't a problem, although I'm not sure what to check it with.[/del] tested with Western Digital Data LifeGuard Diagnostics, extended test came up clean Could stress my CPU for a while, but I don't know if it's really the problem [del]Update chipset drivers[/del] tried that, the installer refuses to work and I double-checked it's for the right chipset [del]Check voltages[/del] tested the 24-pin connector and the pci-e connectors, all voltages checked out fine If you're also having this problem, post in here with as much info as possible and follow these steps above. If anyone can recommend any other steps to take, or possible causes then please post. Also if anyone can infer much from the BSOD log, then take a look at [url=http://pastebin.com/TSrSLRhX]this[/url]. I know that so many people have this issue and I'm hoping plenty of people will post in here. It's probably better to have one centralised base of knowledge which everyone can read rather than lots of occasional threads on the issue which probably won't end up with a fix.
I had this issue with a Nvidia GTS 7600. Apparently my card melted, I had to buy a new one.
I know that in lots of cases it is caused by a card that is dead, I was getting that error before I had to RMA my 570. However, I still have the same problem with my new card and EVGA tell me they thoroughly test all cards they send as replacements so I'd bet against my GPU being the problem here.
Have you tried update your chipset drivers ? Also probably the most important thing, have you checked the power supply voltages ? Another good idea would be to just replace the power supply if you have one laying around, don't worry if it is a little underrated since you'll only be using it for a short time.
[QUOTE=Chryseus;30654311]Have you tried update your chipset drivers ? Also probably the most important thing, have you checked the power supply voltages ? Another good idea would be to just replace the power supply if you have one laying around, don't worry if it is a little underrated since you'll only be using it for a short time.[/QUOTE] I'm running memtest at the moment, when that's done I'll update the chipset drivers. By checking the psu voltages, do you mean manually checking with a multimeter? I don't have a spare psu to test unfortunately.
[QUOTE=halflife_123;30654419]I'm running memtest at the moment, when that's done I'll update the chipset drivers. By checking the psu voltages, do you mean manually checking with a multimeter? I don't have a spare psu to test unfortunately.[/QUOTE] If you have a meter do it manually and measure the AC ripple as well. If not you should be able to read the voltages with [url=http://www.cpuid.com/softwares/hwmonitor.html]this[/url].
[QUOTE=Chryseus;30654490]If you have a meter do it manually and measure the AC ripple as well. If not you should be able to read the voltages with [url=http://www.cpuid.com/softwares/hwmonitor.html]this[/url].[/QUOTE] Unfortunately I don't have a multimeter, I'll see if I can acquire one from somewhere. In the meantime I'll download that and I'll post later once I've done some more testing.
Ok checked my hard drive and it's fine, went to install the chipset drivers from the Intel website and when I ran the installer it stopped and told me I don't have the minimum requirements to install it. I'm not sure why it would say that, I checked in the Device Manager to see which chipset I have and got the correct driver installer (6 series). [editline]23rd June 2011[/editline] Never mind found a different installer. On an unrelated note, Windows Update seems incapable of installing any important updates. [editline]23rd June 2011[/editline] Ok this other installer I found ran fine but didn't actually update the chipset drivers. How handy. On the bright side of things my mobo BIOS is fully up to date. [editline]23rd June 2011[/editline] I ran the System File Checker and apparently it found some corrupt files which it couldn't fix.
Tried thoroughly getting rid of the nvidia sound drivers, which included removing a dll which wasn't covered in the instructions. It didn't crash on startup like it does quite often do so I'll keep my fingers crossed that it helped. [editline]23rd June 2011[/editline] Nope.
Maybe it would be a good idea to do a wipe and reinstall.
[QUOTE=Chryseus;30661673]Maybe it would be a good idea to do a wipe and reinstall.[/QUOTE] Ideally I'd like to do that as a last resort, luckily I have nearly everything on my PC on a external hard drive too so I'll try some more things and if they don't help I'll format and reinstall windows. [editline]24th June 2011[/editline] Here are my voltages, can anyone tell me if there's anything wrong here? [img]http://i86.photobucket.com/albums/k101/halflife_123/voltages.png[/img]
:ms: No wonder you're having problems with the 12V rail being pulled down to 9.15V, assuming the sensor is correct then the power supply is definitely the cause of your problem. It probably would be a good idea to get a cheap £5 meter just to check it manually before buying a new supply.
[QUOTE=Chryseus;30686295]:ms: No wonder you're having problems with the 12V rail being pulled down to 9.15V, assuming the sensor is correct then the power supply is definitely the cause of your problem. It probably would be a good idea to get a cheap £5 meter just to check it manually before buying a new supply.[/QUOTE] I'll look around for a multimeter, thanks. Will a low voltage actually cause any long term damage to anything?
[QUOTE=halflife_123;30687403]I'll look around for a multimeter, thanks. Will a low voltage actually cause any long term damage to anything?[/QUOTE] Low voltage will not cause any damage.
Ok, got myself a multimeter and tested the 24-pin power connector. The 12v wires actually had voltage of 12v (not 9.xx like HM was reading), out of interest I tested the 3.3v and 5v and they were 3.37v and 5.05v so I guess that's enough proof that the software readings were wrong, and that my PSU probably isn't a problem unless there's something else about it we've missed.
I used to have this problem, I'm not really sure what I did to fix it though. My driver would crash randomly and recover and just keep going. I remember I go and push down on the graphics card and make sure it was in right and it would seem to go away for awhile.
[QUOTE=halflife_123;30766314]Ok, got myself a multimeter and tested the 24-pin power connector. The 12v wires actually had voltage of 12v (not 9.xx like HM was reading), out of interest I tested the 3.3v and 5v and they were 3.37v and 5.05v so I guess that's enough proof that the software readings were wrong, and that my PSU probably isn't a problem unless there's something else about it we've missed.[/QUOTE] You tested the 24 pin connector.. so the computer was not actually running ? Also depending on the supply the 12V rail can be split so you need to measure the 12V going to the GPU. While it is running put a game on as well to increase the GPU load. If that all checks out you should do a reinstall of the OS.
[QUOTE=Chryseus;30767909]You tested the 24 pin connector.. so the computer was not actually running ? Also depending on the supply the 12V rail can be split so you need to measure the 12V going to the GPU. While it is running put a game on as well to increase the GPU load. If that all checks out you should do a reinstall of the OS.[/QUOTE] The PC was running, how else could I have tested it? I'll test the GPU voltages tomorrow and post again.
I have this same issue, but recently I got even more on top of it. My computer started to lock up randomly, the LCD poster that's on it just said "USBfinal". After some searching I found that it might be related to RAM. And some more time passed after that I got BSODs that read "Page_falt_in_nonpaged_area" and that's also a RAM issue. Also, the video card crashing due to TDR can also be a RAM problem as I have read. So I ran Memtest on each one of my 3 sticks of ram (2GBx3) each with 1 pass, then all 3 with 2 passes. It came up with 0 errors. After installing all 3 of them I didn't have the problem for about three days, It started up again though. Anyway, perhaps you should eliminate one stick of RAM from your computer and run it like that for some time, and if it happens again, switch sticks and repeat.
Ran the DiRT 3 benchmark test while I checked the GPU voltages, they never dropped below 11.95v. [QUOTE=Shock_Coil;30772884]I have this same issue, but recently I got even more on top of it. My computer started to lock up randomly, the LCD poster that's on it just said "USBfinal". After some searching I found that it might be related to RAM. And some more time passed after that I got BSODs that read "Page_falt_in_nonpaged_area" and that's also a RAM issue. Also, the video card crashing due to TDR can also be a RAM problem as I have read. So I ran Memtest on each one of my 3 sticks of ram (2GBx3) each with 1 pass, then all 3 with 2 passes. It came up with 0 errors. After installing all 3 of them I didn't have the problem for about three days, It started up again though. Anyway, perhaps you should eliminate one stick of RAM from your computer and run it like that for some time, and if it happens again, switch sticks and repeat.[/QUOTE] I still haven't ran memtest for a long time, longest I left it was 2 hours. Strange though, it took 40 mins to complete the first pass but after 2 hours it hadn't done another even though the progress bar was still increasing, I just assumed it was a bug with memtest. I'll try running memtest with each stick individually and then all together and see what happens.
Ok ran each 2GB module through memtest on their own with 2 passes, then put all 4 in and got 4 passes. I may try taking one out and running my PC for a while in case the test didn't bring up a possible problem, but I also have other options to consider. One peculiar thing I noticed was that when I took out a module and replaced it with another, there would be no video output each time. I checked the module was in properly and nothing changed, I checked my card was still seated properly and tried turning it on again and it was ok. I'm thinking that I was slightly moving the card or the power connectors when I was changing the RAM modules, and now I'm thinking that perhaps this may possibly be the cause of the problem. Strange though, I checked the voltages one time when it booted up with no video output and they were fine, so perhaps even the slightest movement might be taking it out of the PCI-E slot slightly or something. This was peculiar. [editline]1st July 2011[/editline] Little update here: anticipating the worst, I put in a SSD which I had completely forgot about when I assembled the PC and decided I would back-up anything I already hadn't, but when windows loaded it crashed and tried to recover it loads of times, so I just formatted my hard drive and put windows on the SSD and put all my stuff on. Everything seems ok at the moment.
I might have had a break through here. I was running my computer on one stick for the entire day, other then me having to tune down TF2, it was all fine. Today I replace stick 1 with stick 2, I loaded up a youtube video and my screen flashes as soon as it starts playing and I get the error. I shut down the computer and replace stick 2 with stick 3. I load up the same video and the same thing happens, Then I put stick 1 back in, play the same video and no crashing. So it's either two bad sticks or just another coincidence. I really think that you should do the same test as I did. If anything just run your computer for a day on each stick.
[QUOTE=Shock_Coil;30825911]I might have had a break through here. I was running my computer on one stick for the entire day, other then me having to tune down TF2, it was all fine. Today I replace stick 1 with stick 2, I loaded up a youtube video and my screen flashes as soon as it starts playing and I get the error. I shut down the computer and replace stick 2 with stick 3. I load up the same video and the same thing happens, Then I put stick 1 back in, play the same video and no crashing. So it's either two bad sticks or just another coincidence. I really think that you should do the same test as I did. If anything just run your computer for a day on each stick.[/QUOTE] I'm going to go play some games, so far mine has been ok but I haven't really done a lot to make it crash - only an hour of TF2. It crashes more often in World of Tanks so I'll play that for a while and see how it goes. If I still get crashes I'll take some RAM out. The only trouble is there is no definite way to make it crash (at least in my case) so it could be difficult to work out if I have a bad stick or not. [editline]1st July 2011[/editline] Ok so it happened a few times during WoT, it was different to usual though. When it used to happen the screen would go black and my monitor would tell me there was no input, then it would come back. Now it goes off and comes back almost instantly with no message. I might take some RAM out now.
That's the same thing that happens to me. So, I was just playing TF2 with stick 1 in my computer and guess what happened, black screen. So either all of my sticks are bad or my computer just loves to screw with me. I'm not out of conclusive tests, nothing seems to be the problem yet everything seems to be the problem. I might try to run with only one card at a time and just rotate it with the three I have.
Chiming in here, same issue. Screen goes No Signal, sometimes recovering other times (mostly) not. My PC's open on my bench right now and it seems to be working, again. Beforehand I couldn't to the desktop. Seems to alternate between periods where nearly anything will black screen and others where furmark runs fine.
Same, mine can be fine for hours then do it every 5 minutes which is annoying. I'm still trying to figure out if my RAM is the culprit, I'm wondering how tiny the chances are of having 2 bad sticks compared to another problem I haven't thought of yet...
I get this every time I turn on my external hard drive. Yeah.
Well I tried sticks 1 & 2 and had a crash, then 3 & 4 and had a crash. Tried 1 and 3 and I've had no crashes today, I'm not sure whether to hope that I have 2 bad sticks or that it's just sheer luck, because if they aren't duds then it means it's something else and I'm running out of possible causes... [editline]5th July 2011[/editline] Ok it must be the RAM, I've had no issues since I put in sticks 1 & 3 on their own. I dunno how to go about returning it or what I should do really, since I need it and it could be a while before I get some new RAM. They might even want to test it and see if it is damaged, although I hope not since memtest and general PC usage don't cause any crashes and they may very well just send me the same stuff back telling me it works fine...
Well quite a big bump here, I have 3 sticks in now and it works fine. I'm now thinking it has something to do with having all the RAM in dual-channel mode or something, I really have no idea though.
It sounds very possible that ram could be the problem. Ever since i installed linux i've never had a crash; I'm guessing linux allocates ram to programs differently than windows. Tell us if you have a crash again, i would love to know if that really was the problem.
Sorry, you need to Log In to post a reply to this thread.