175.16 - NVLDDMKM.dll error still happening... anyone else?

Status
Not open for further replies.
Problems went away when I got to 600 Core for my GTX. I am sure its losy cooling fan on the card. I put an aftermarket cooler on the card. Has improved the cooling. I bought 8800 gtx XXX XFX and its suppose to work at 630 but not even close. I bought Zalman fan and chipset cooler and its much better.

Its overheating issue that caused my card to crash in 3D games.

Stock thermal pads just slipped off with no effort. Bad quality control.
 
Got this off another forum :

NVIDIA statement on TDR Error Messages

Some Windows Vista users have reported that their systems are displaying an error message that says: "Display driver stopped responding, but has successfully recovered." This is called a Timeout Detection and Recovery error message.

Timeout Detection and Recovery (TDR) is a new feature of Windows Vista that attempts to detect problematic situations and recover to a functional desktop without forcing a reboot. Hangs can occur when the GPU is processing intensive graphics operations, typically during gameplay, and nothing is being updated on the monitor. To the user it appears that the system is frozen with no resolution to the problem; in previous operating systems users generally had to wait a few seconds and then reboot.

The TDR error message "Display driver stopped responding and has recovered" lets the user know that the NVIDIA display driver (specifically the "nvlddmkm.sys" file) has been re-initialized and the GPU is reset without requiring a reboot. The only visible artifact from the recovery is a screen flicker, the result of a screen redraw. Note that some older Microsoft DirectX applications may render to a black screen at the end of the TDR, requiring the user to restart these applications.

TDRs are not specific to a single driver problem, and can occur for a variety of reasons. When they occur, diagnostic information is collected in the form of a debug report that is sent to Microsoft through the Online Crash Analysis (OCA) mechanism if the user opts to provide feedback.

NVIDIA encourages users to submit their own bug reports via the NVIDIA Vista Quality Assurance Program, using the keyword "TDR" in the description of the problem. The NVIDIA bug report link is here:
https://surveys.nvidia.com/index.jsp?pi=7498eac864dc1950c8f09e040b4a437a

We understand that many users have expressed frustration with this issue, and we apologize for the inconvenience. Since the NVIDIA v101.41 beta driver release, NVIDIA has been fixing many TDR issues reported by users. Our software team is currently preparing a new driver which will dramatically reduce the number of TDR errors that users have reported on the forums. Thank you for your patience.

More information on TDRs can be found here:
http://www.microsoft.com/whdc/device/display/wddm_timeout.mspx


..Go submit the issues to Nvidia and let them know you're having this problem!

EDIT : forgot to mention that just today I started having this issue again. I can run all my games just fine from dx9 to dx10 games. This is now happening when I am updating my website (uploading files and such) and when I have Photoshop cs3 open on my main monitor. This has happened 3 times in the last 6 hours I have been home.

I've been on many forums, and I'm trying out some new ideas.
I will get back to you guys to let you know if anything seemed to work.
I have not had this issue in over 6 months since replacing a stick of ram until this morning.
Seems like this forum screwed me. :eek:
 
The TDR excuse is the most likely culprit since this happens to ATI cards as well. But hey, who knows.
 
Ok, so I've given up on eVGA entirely and have just contacted NVIDIA's engineers directly. I'll let you know how that goes.
 
Ok, so I did get an answer from our NVIDIA reps.

"WinVista changed the way a device driver interacts with the operating system. Pre-Vista, the driver had direct access to Kernel mode functions and if a driver misbehaved, the whole system would crash. With WinVista, all non-MSFT device drivers are moved out of the kernel and into the user mode space so a bad driver will not crash the system. Instead, the system will load a generic MSFT driver, reset the 3rd-party driver and reload it all without crashing the system. The NVLDDMKM error is what the user sees when a TDR has occurred and the driver has reset.

This only happens in WinVista. WinXP will crash if there is a driver hang.
TDR is the side-effect or result of a driver/application issue.
Newer drivers may help resolve the number of TDR’s.

There is no way to get the TDR’s to stop without figuring out what is causing the original driver faults. You really need to treat this as any other standard tech support issue and have your techs do what they can do to help identify the root cause. Bad hardware or programs can also force a video card TDR.

More and more users are adopting WinVista which is why you are encountering more cases now. If these cases are happening too frequently, report this to your card vendor partner. They have the ability to log these issues with us directly."

So, it looks like the best that all of us can do is to report the issue directly to the card manufacturer and let them interact with NVIDIA.

If you guys need any help, I can try to get you contact information.

Heather
 
So basically something is wrong and causes the TDR, and since all we see is the TDR we think that's what the issue is, when in fact it's the side effect. Makes sense given the fact that it appears to be resolved or unresolved due to lots of different changes.
 
Yeah, that explanation was the most straightforward one I've ever gotten, and it makes a lot of sense...and it also explains why different things work for different people who are having the same error.

Now that we really understand the WHY, maybe those who are having issues can contact the card vendors and get some help.
 
Heather Taylor said:
*snip* Newer drivers may help resolve the number of TDR’s. *snip*

Heather
Click to expand...

My question still is, which newer driver?? I've been trying drivers left and right, and haven't found one that is making this problem stop. Very frustrating.
 
BigDov said:
My question still is, which newer driver?? I've been trying drivers left and right, and haven't found one that is making this problem stop. Very frustrating.
Click to expand...

Yeah, I feel your pain, and I think that what you've written here is the absolute root of why this error is such a nightmare. There's no one fix for this.

What Nvidia is basically saying is that something in the system is wrong, which causes the TDR to fail, which then causes the driver error. We were all thinking that it was the driver breaking that caused the TDR to fail, but we had it backwards.

So, somewhere in your system, there is another, deeper problem which drivers aren't fixing for you. Without seeing your specs, I can't tell you if there's any other hardware conflicts I can identify straight away.

If I were to treat this like any other technical support issue, I would start you out by having you run Memtest just to make sure your memory is ok. Like many on here have said, that can be one cause of the problem. If your memory passes, I would tell you to contact the manufacturer of the video card so that they can log the issue with NVIDIA. You might just have a bad video card.
 
Just as an FYI I ran the vista memory diagnostic and it did report RAM issues. So if I get those squared away then hopefully that's my particular TDR issue.
 
Check to make sure your timings are right on your memory. Incorrect timings can give you that TDR stuff, and it can also just cause some general havoc.
 
Heather Taylor said:
Yeah, I feel your pain, and I think that what you've written here is the absolute root of why this error is such a nightmare. There's no one fix for this.

What Nvidia is basically saying is that something in the system is wrong, which causes the TDR to fail, which then causes the driver error. We were all thinking that it was the driver breaking that caused the TDR to fail, but we had it backwards.

So, somewhere in your system, there is another, deeper problem which drivers aren't fixing for you. Without seeing your specs, I can't tell you if there's any other hardware conflicts I can identify straight away.

If I were to treat this like any other technical support issue, I would start you out by having you run Memtest just to make sure your memory is ok. Like many on here have said, that can be one cause of the problem. If your memory passes, I would tell you to contact the manufacturer of the video card so that they can log the issue with NVIDIA. You might just have a bad video card.
Click to expand...

I'm going to guess it's the video card. I didn't have this issue with the last video card at all- I'm going to try one last driver wipe and then give up.

Specs though, I almost cringe to list......but here goes:

2x 2222 Opterons
8x2gb dimms
9800GTX
3x500 hdd
1x146gb sas drive w/ LSI controller
1000w psu
oem branded board- it's a Sun Ultra 40M2 workstation

Memory has passed TONS of memtest since I've had this, and I've only had these problems since installing the 9800, which is unfortunate. When it works, it works excellently.

I really appreciate the info and input you've had here Heather. Very cool.
 
You may be right, it may be the card--or it may be the card along with the memory timings. Your guess is as good as mine there. While your specs aren't really cringe-worthy, they are definitely unusual for this forum.

Since it's an OEM branded board, is that memory ECC? That may create an issue. Memtest has problems when testing ECC memory.

Also...That's a lot of memory you've got there, my friend. Wow.
 
Heather Taylor said:
You may be right, it may be the card--or it may be the card along with the memory timings. Your guess is as good as mine there. While your specs aren't really cringe-worthy, they are definitely unusual for this forum.

Since it's an OEM branded board, is that memory ECC? That may create an issue. Memtest has problems when testing ECC memory.

Also...That's a lot of memory you've got there, my friend. Wow.
Click to expand...

:) They are a bit unusual, you're right. I've kind of gone away from tweaking everything to just having something that works, and works well for all of the video editing, photoshop, etc that I do.

It is ECC, but I never saw any errors. I'll launch it tonight before going to bed and let it burn away, and see what I wake up to tomorrow.

I know..... it's silly really. Vista 64 loves it though!


edit- I meant to say, I did the driver clean in safe mode and all of that, re-installed the 175.16's and haven't had any problems yet. The games that would normally be causing the dump to the desktop, aren't dumping at the moment *shrug* I don't get it!!!
 
Are you running the memory at 1T command rate? Just curious. That's usually the preferred setting for ECC.

If you've had the issue since getting your video card, you could try what Ryan did and underclock the card slightly--again, this would be a good thing to run by the card manufacturer, if you can get someone in 2nd level support.
 
Hey, just wanted to add my two cents, explain my story, and bump this thread, I apologize for the huge wall of text. I purchased a dell XPS720(big mistake already, I know) just a few months ago, and began using it to my leisure with little to no problems at all. Upon powering up my computer for the first time, I merely went with the files that were already installed and downloaded a few window's updates and began playing the games I normally play(World of warcraft, warcraft 3, counterstrike, DoD/CS source, and Crysis). I would experience a BSOD maybe every 2-3 weeks and was fine with that, I didn't expect my machine to be perfect. So a few weeks after owning my dell I began seeing the nvlddmkm.sys error which I assumed was caused by a windows update due to the fact that I have NEVER overclocked my PC or downloaded anything out of the ordinary.

So untill that point I had been using the video card drivers that were available to me via windows update, and I decided to download the most recent set from nvidia.com(driver 175.16). After this download, my computer went to shit. I couldn't play any game for more than a minute without having the game crash resulting in a nlvddmkm.sys error. So I performed a system restore, and was STILL having this error. Tried going back to the original graphics driver that was available to me through window's update and I was still getting the problem. I too have tried EVERY attempt at resolving this problem as the original poster has, nothing has worked or even slightly fixed the problem. I've reformated on multiple occasions where I've downloading nothing but SP1 and the newest drivers, I've even tried the window's update driver and that still causes the error. I've tried removing a stick of ram.. Still seeing the error. Tried Stardock's Tweakvista.. Still see the error.

I have just reformated and am waiting for someone to possibly sit down with me and hopefully try to aid me in fixing this terrible issue. At the moment I have vista 32bit, only service pack 1 installed, and 14 copies of nlvddmkm.sys in the system32 folder





http://img386.imageshack.us/img386/3268/nlvdfolder2xl6.jpg
Files#2^

http://img168.imageshack.us/img168/4089/nlvdfolder3mj7.jpg
Files#3^

http://img72.imageshack.us/img72/8933/nlvdfolder4wz4.jpg
Files#4^


I made a post regarding the matter right here:
http://www.ocforums.com/showpost.php?p=5664924&postcount=288
If you could check it out, I'd greatly appreciate it, if you require any additional info please let me know.(my computer spec's are in the thread I linked above)
THANK YOU!
-Sincerely, Sean
 
Sean,

Let me see what I can come up with. The fact that you were having bluescreens randomly before this error started concerns me--while no PC is perfect, it certainly shouldn't be bluescreening like that unless there's a problem.

For now, start by running Memtest 86 for at least three passes--which means you should probably start it before you go to bed and let it go all night, since each pass can take over an hour. Removing a stick of RAM can sometimes identify one that's bad, other times, the problems with memory only show up intermittently and removing a stick is inconclusive.

Let me know what the results are and we'll take it from there.
 
I've been doing testing over the past month or so, and here's what I've come up with so far:

SHORT VERSION
The root of the problem is pop-ups in DX10 games, such as windows update notifications, instant-messaging programs like trillian or msn or ICQ etc.

LONG VERSION
The crashes only occur when a pop-up window occurs while running in DX10. The crash will not happen in DX9 games.

So, for example all the following will cause the NVLDDMKM.sys/TDR driver crash:
- a volume increase / decrease button on the keyboard, which pops up a tiny window showing the new volume
- windows update notification pop up
- low battery mouse warning pop up
- instant messaging programs

I can repeat these results without fail in any DX10 games. As for hellgate london crash instantly whenever you change the graphics settings and enable dx10, I'm willing to bet thats just a poorly coded game. No other DX10 games such as crysis or World in Conflict have that problem, they ONLY crash when a pop up occurs.

For all the naysayers that it's memroy or some junk like that... I also replaced my memory with 4x2gb OCZ reaper PC8500 ram just for the sport of it. Also, I went to vista64. No difference. Also, running at stock clocks on everything, nothing is overclocked in the system whatsoever.

For now, I've basically eliminated the crashing by simply turning off anything that can pop-up while running a game. It's not the most elegant solution, since it should work no matter what, but at least I can play games now.
 
hows this for simple info....

my system in my sig, runs UT3 beautifully in SLI, Vista x64 SP1 fully updated....175.16 NV driver loaded

i changed one single thing, that was to go into the Nvidia control panel and edit the SLI setting for the UT3.exe executeable, i changed it from SLI(NVIDIA RECOMMENDED) to force AFR2

launch UT3, and during loading the beginning of the game (actually launching it, past the menus) it crashed and i got the BSOD with the NVLDDMKM.dll error

rebooted, figured it was a fluke, launched UT3 again, same exact thing happened

changed SLI setting back to the default as it was for UT3.exe, launched game, no issues since

so there you have an example of this coming up with only one single change being made to my system, and reversing that change fixed the issue on this system ;)
 
Guys,

Thank you both for this info. I've passed it up to NVIDIA, and I'll be curious to see what they have to say, at least about Ryan's issue.
 
I just loaded Vista x64 yesterday. I'm getting the same error everyone else is getting when I log into a server in CoD4. The screen freezes for a minute when I first start on a map, then its good to go after that. Running two 8800GTS 512 in SLI and 8GB of RAM. Other than this issue, I'm grooving on my new OS.
 
Didn't have any problems with it once I disabled SLI. That's cool, but it suxors in a way. Looks like I'll be hitting up the EVGA step-up to a 260 for the newer one and shove the other one in my back-up box.
 
That's what I suspected--that is one thing that I have noticed. You may not have to stick those cards in a box somewhere...the secondary card may be bad, or even the SLI bridge might be messed up.

If you've a mind, take the cards out, then put the bottom card in the top slot. If that second card is bad, you should get the same error when you play with one card. If you can play smoothly, then ask eVGA to send you a replacement bridge and try that.
 
I would like to chime in and tell you what I have found. I am currently building a new PC:

MSI P7n SLI Platinum
E6600 with Asus 80mm cool with artic silver
PNY 9800 GX2
Gskill 2Gx2
BFG 650watt power
I have removed/disconnected everything else that’s not needed to troubleshoot this error.

Loaded Vista Premium x64. Installed SP1 and all Windows updates. Installed the newest nForce and GeForce drivers from Nvidia. All seemed fine. Ran 3dMark05 to test setup and got blue screen “NVLDDMKM.sys Attempt to reset the display driver and recover from time out failed”

Ran Driver Cleaner Pro and proceded to try every driver I could find. I used Driver Cleaner Pro after each try.
GeForce 175.16, 174.74, 174.53, 175.16, 174.74
nForce 15.17, 15.08
Then tried the drivers from MSI’s website for the P7n SLI Platinum.
Tried different combination of drivers.
Same error everytime.

Verified BIOS was current and even tried a beta Bios. Set RAM timing’s as recommended 5-5-5-15. RAM voltage recommends 1.8-1.9 tried both.

BIOS temps looked fine. Installed SpeedFan and temps look good there.
CPU 33c
Vid 60c

Took the RAM out of my current PC and tried it in the new PC. (Same ram Gskill 2Gx2) and I tired another new set of the same type of RAM. Still getting error when starting
3dMark05.

Decided to try Vista x32. Installed all patches and SP1. Same error again. Tried different driver combinations each ending in total failure. PC works fine until anything graphic intensive is started.

Formatted once again and installed XP Pro x32. Patched it all up. Loaded the newest drivers started up 3dMark05 and got Blue Screen! This time it was for N44_Disp. Tried different drivers and got same result.

Pulled the PNY 8800GTX out of my other PC. Formated hard drive and installed XP pro again. Did all the updates. Newest drivers. Started 3dmark05 and it worked fine. No errors to be had.

Bad video card? After spending my whole weekend troubleshooting this I am wondering if that’s all it is. Whats the difference in the newest drivers for the 8800GTX and the 9800 GX2? Is there a difference? The both have the same version number.

Update:
Just formated and installed Vista x64 with my 8800GTX installed. Install kept crashing at the very end with a blue screen. Too quick to read and Windows never loaded.

Performed a memory test using Windows Vista DVD and got "Hardware problems were detected". Did not say what it found though. Removed a stick of RAM and Vista loaded no issues. Put the RAM back in once install was done and all is well.

I then installed just the GeForce drivers 175.16. Did not update anything or install SP1. Ran 3dMark05 and got 13955 and no blue screen.

Put the RAM into this computer (one in the sig) and ran the same memory test. Nothing found.

I'm now leaning towards RMAing the motherboard. I was going to place the 9800GX2 into this computer but this power supply does not have a spare 8 pin adaptor and I really don't feel like taking everything apart tonight. I have it looking so pretty and clean! Maybe after work tomorrow I'll try it.
 
Damn Poore, that is some serious troubleshooting. I know when I was running Win XP Pro with the same setup I have now (minus 4GB of mem), I had no issues. According to all the information available, its a Vista issue only. With mine, it happens sometimes when I first log into a server. It won't crash back to the desktop, but it does freeze for a second or two. I'm going to take out my older GTS 8800 512 and drop it into my backup gaming rig. Then my newer one is going back to EVGA for a stepup to a 260. Can't say its going to stop the problem, but I've had nothing but issues every since I went to SLI in this rig.

p.s. Thanks Heather for your help. Seems nVidia and Microsoft need to work together and try and figure this mess out. I know nVidia is saying its a hardware issue that is at the root of the problem, but then that insuates that we all have shitty hardware because there is a hell of alot of people on the net bitching about this issue. I think nVidia has a software programming issue, and they can't figure out how to fix it.
 
The same error basically happens with ATI cards, too. The best I can tell is that it's Vista more than it's NVIDIA...which doesn't mean that NVIDIA shouldn't fix the issue, just that it's not limited to their cards.

Poore, try 175.19, too. It could be the motherboard, but it could also be the video card...it's hard to tell at this point.
 
Tonight I am going to place the 9800 in my good PC and see what happens. If I have issues I'll RMA the video card. If it works fine the system board is getting RMA'ed. This is the only thing I have not tried yet.
 
does anyone else here get crashes when pop-ups occur with SLI during DX10 games? Thats the ONLY thing that causes my crashes that I can see so far. Doesn't happen in dx9 games. Vista32/64 doesnt matter, both have the problem.
 
I just heard back from Nvidia, Ryan. I gave them all of the original info at the beginning of this thread, as well as the bit about pop-ups, and their lead engineer is researching it for me. Does the error happen outside of SLI? (You may have said so before, but it's buried in this thread somewhere.)
 
I'm still having the same issue after having my 8800GTX & surge protector replaced. I've whittled the issue down to drivers, GPU, RAM, and/or PSU. At the moment, I'm waiting for OCZ to RMA my RAM. If the replaced RAM doesn't fix it... I'm down to mucking with a new vid card (4870, 4870 X2, or GTX 280) or a new PSU. There's way too many variables through everyone's issues to be one exact thing. TDR might be too touchy... or certain drivers might have a horrible implementation of certain Vista API calls. I don't have a clue at the moment & I've been dealing with this since February 2007. It's sad when neither Microsoft, AMD, Intel or Nvidia can provide a straight answer to any of our issues. It gives a bad name to PCs as a whole if I can't play games as I'd like on my highly favored platform. I'm not really into consoles & I don't want to waste what I've invested in my current PC (which is quite new - PSU is the only "old" component). Will this crap every end?!
 
It's definitely TDR related. That much we know. It's the THING that triggers the TDR fail that's the unknown variable.
 
Can anyone confirm that this issue goes away with a new GTX 260 or 280? I'm thinking this issue started kicking in around the time G80 GPUs were released and has gone well into the G92s. If this is the case, I'll gladly buy & test out a new GTX 260 to sate my hunger for crash-free gaming from here on out.
 
If memory serves, I started seeing these crashes after the emergence of the 7900GTX cards paired with Vista.
 
I found something that seemed to fix on my end - seems very stupid, but I've been able to model in XSI without a crash for ~15min now, where I'd crash in ~30sec before.

In the bios clock settings, I set my PCI-E clock to 100mhz manually rather than auto. Should be the same, but I'm much more stable now.

Gigabyte P35-DS3L 2.0
E8400
eVGA 8800GT
4gb A-Data RAM PC800

Seems to have done the trick - now that I've posted however, surely it will break in just a moment.

edit: And as I typed, problem rears its head the SECOND I start playing with it again. Alas.
edit2: I am utmost confident that it IS tied to my dual monitor setup. When running just one, I'm stable as a rock for hours+ however I need the second one for references / documentation :(
 
If you're having the error running two monitors, I wonder if one of your DVI ports has a problem? Contact eVGA and see what they have to say.
 
Heather: emailed ya back. As for does the pop-up cause crashes while NOT in sli... I actually cant recall testing this. I have to give it a shot and get back to you.

I do know Hellgate London insta-crashed with even 1 card though when you enabled DX10 / changed to 2560x1600 etc.
 
I updated to the lastest patch in CoD4 today. Wouldn't even let me log into the server. Kept kicking me out with DirectX errors and the NVD blah blah blah error. So i took one my cards out and killed the whole SLI deal, and now I've got no problems. Works like a champ. I'll be doing the stepup to the 260 immediately. What a goofball issue this is.
 
I checked my Vista administrator logs after my system crashed twice in a row running 3d apps. I've been getting nvldmkm critical errors for the past month and just now started having issues. I rolled back to 175.70 a couple days ago and haven't had 1 issue yet.
 
Status
Not open for further replies.
Back
Top