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

Status
Not open for further replies.
TigerLord said:
It did for me twice.

Plus try Google. The OP tried every software solutions.

You have another solution to help the OP?
Click to expand...

Your exact statement was "this is often a ram issue." And that's what I disagree with.

Now speaking logically, I posted exactly what I did to fix the problem on my system. How would I post a second solution? Undo my fix and see what happens?

That makes no sense.
 
I'm saying DX10 games in Windows Vista are the only problem. I can play DX9 games like warcraft 3 or supreme commander forged alliance for days on end without a single crash.

Also, just for the sport of it, I JUST finished another run through of memtest86 3.4a. No errors. People claiming it's the ram... it isn't.

I'm making this thread because I'm sick and tired of companies claiming this crash is caused by user error or is merely a random occurrence that they have no effect on, and consequently bear no responsibility for the crash. It isn't entirely random, this crash is occurring because someone's drivers or software is messed up. Period. A system that has been working perfectly fine - and continues to do so in windows xp - does not mysteriously cease to work only while in windows vista while playing DX10 games.

This is software related, whether it be the OS, the software, or the software driving the hardware. I'm trying to get this thread to get a little bit of attention so maybe we can bring it to the attention of Kyle at the [H] and maybe get a little front page loving.. perhaps put a little heat under nvidia, they might respond if people start demanding answers... Oh, and I'm writing this right now while running all 4 cores at 100% maximum with prime95. I'm running the test again to please the nay-sayers, but I know this is going to come up negative as well. The system won't crash. I use this machine daily for my business, and it's been otherwise solid as a rock. I've run these tests many times before, and there was never a problem then, and I'm not surprised they're coming up negative again today with zero problems.
 
I was having this problem. I did what g0at suggested and uninstalled the drivers. used the microsoft ones, then I decided i didnt want the microsoft drivers, so i reinstalled the nvidia drivers.

this is different then how i installed the nvidia drivers before, i just installed the nvidia drivers upon installation of windows and never used the microsoft drivers from update.

i havent been getting any more CTDs or BSOD as a result of the nvlddmkm
 
Tetrahedron: I wasn't able to do exactly what you said. I uninstalled the 175.16 WHQL drivers, rebooted, and ran windows update which detected a new version of the driver for 8800gtx. I attempted to install it twice, but it gave the same error both times, and the error code basically described that it would not allow it to be installed because it was older than the driver currently existing in the system (even though the nvidia driver was uninstalled). So I manually went into device manager and selected the 8800gtx driver that is installed by default for both cards. I have now installed 175.16 over top of this driver, without uninstalling it.

We'll see what happens.

(P.S. the default windows driver is forceware 163.69)
 
If you notice one thing, and the one thing that is causing all the errors, is a conflict between the creative drivers and nvidia drivers.

Creative, Vista, and Nvidia are all to blame, but if you want to get rid of this error forever, ditch your creative soundcard and all its drivers and I assure you it will go away.

Worked for me, and I have been noticing that basically everyone with this error either has a Creative Sound Card, or just having RAM/PSU issues.
 
Talnayen said:
If you notice one thing, and the one thing that is causing all the errors, is a conflict between the creative drivers and nvidia drivers.

Creative, Vista, and Nvidia are all to blame, but if you want to get rid of this error forever, ditch your creative soundcard and all its drivers and I assure you it will go away.

Worked for me, and I have been noticing that basically everyone with this error either has a Creative Sound Card, or just having RAM/PSU issues.
Click to expand...

i use onboard audio, realtek, since like 2004, not an audiophile, just use headphones myself
 
Talnayen said:
If you notice one thing, and the one thing that is causing all the errors, is a conflict between the creative drivers and nvidia drivers.

Creative, Vista, and Nvidia are all to blame, but if you want to get rid of this error forever, ditch your creative soundcard and all its drivers and I assure you it will go away.

Worked for me, and I have been noticing that basically everyone with this error either has a Creative Sound Card, or just having RAM/PSU issues.
Click to expand...

I'm using a Razer sound card.

Though I have read some people thinking that it's a realtek conflict. My onboard sound (unused) is a realtek chip.
 
I have removed my X-Fi elite pro and am now running onboard sound only (680i), I installed the latest realtek drivers - the R195 series. We'll see how this goes.
 
I have an 8800GTX, 780i mobo (w/ latest BIOS version from EVGA that corrects graphic corruption), and the Auzen X-Fi Prelude on Vista Ultimate x64. This issue has occurred with every driver version with 163.69 being the most stable but it still crashes. I'm starting to wonder if it is a Creative, Nvidia, & Vista conflict... or an issue with G80 GPUs and above on certain systems. I'm still considering that I'll buy the GTX 280 & attempt to see if anything improves.

I have also tested this on XP x64 SP2 in recent times & the issue still occurs.
 
I've seen it with people using onboard sound, not a Creative card, so that kind of takes Creative drivers/hardware out of the mix. I've also seen it on motherboards dating back to 939 chipsets (my own system at VM had this problem, and it was an A8N-SLI premium board). At its very root, this is an Nvidia problem...the rest of the variables change depending on the build.
 
I was researching the Nvidia forums & found that increasing the PCI-E speed to 110 can resolve some of the issues. Might have to do with some strange bottleneck. At the moment, it's working well for me.
 
Another thing you might try is a BIOS update, if possible. If you haven't tried it yet. Worked for some folks.
 
FYI - don't OC the 3rd PCI-E slot on a 780i mobo. It will possibly crap out.

Continuing to test out different methods.
 
My friend is having this problem right now with his brand new computer that we built only a week ago. First we noticed his sound was crapping out when we watched videos. Then the error messages as in the OP occurred during AoC. Now they're happening a lot more often. I don't know if it's related to the sound issues.

He's using:
HT OMEGA STRIKER 7.1 Channels PCI Interface Sound Card
ECS N8800GTS-512MX GeForce 8800GTS (G92) 512MB 256-bit GDDR3 PCI Express 2.0 x16 HDCP Ready SLI Supported Video Card x2
EVGA 123-YW-E175-A1 LGA 775 NVIDIA nForce 750i FTW SLI ATX Intel Motherboard
Q6600, 4 GB RAM, Vista...
It's not overheating, thanks to the awesome heatsink we got it's only going to 50s at most. Stays in 40s regularly and he doesn't overclock.
He has drivers all downloaded during the past week, including his BIOS.

Folks at the AoC board are saying it's NVidia SLI + Vista that's causing the problem and that it's best to disable one card. But that's imo not a solution since he bought two cards just for that game.

However using only one of the cards does work....the errors don't continue with SLI disabled. :X
 
It also happens with single card setups.

I've often wondered if it's something in the card BIOS' that create the issue...which brings us back to clock speeds, etc.
 
RyanLucier said:
I'm saying DX10 games in Windows Vista are the only problem. I can play DX9 games like warcraft 3 or supreme commander forged alliance for days on end without a single crash.

Also, just for the sport of it, I JUST finished another run through of memtest86 3.4a. No errors. People claiming it's the ram... it isn't.

I'm making this thread because I'm sick and tired of companies claiming this crash is caused by user error or is merely a random occurrence that they have no effect on, and consequently bear no responsibility for the crash. It isn't entirely random, this crash is occurring because someone's drivers or software is messed up. Period. A system that has been working perfectly fine - and continues to do so in windows xp - does not mysteriously cease to work only while in windows vista while playing DX10 games.

This is software related, whether it be the OS, the software, or the software driving the hardware. I'm trying to get this thread to get a little bit of attention so maybe we can bring it to the attention of Kyle at the [H] and maybe get a little front page loving.. perhaps put a little heat under nvidia, they might respond if people start demanding answers... Oh, and I'm writing this right now while running all 4 cores at 100% maximum with prime95. I'm running the test again to please the nay-sayers, but I know this is going to come up negative as well. The system won't crash. I use this machine daily for my business, and it's been otherwise solid as a rock. I've run these tests many times before, and there was never a problem then, and I'm not surprised they're coming up negative again today with zero problems.
Click to expand...


I'd honestly say that ruling out anything right now is a bit crazy. I started out with 4gb of ram right off a fresh install of Vista x64. I started having the issue without even trying to play a single game. Removed that stick of ram that memtest86+ said was fine, and the problem went away. Put that stick back in, and the error started back up. Either way, it's got to be a combination of different things (drivers, dx10, ram, bad something or other that is possibly causing this issue, or maybe alot of different things cause this issue. Again, this is Microsoft only problem, so we don't know unless we test everything we possibly can.
-g0at
 
TigerLord said:
It's been known for a while the this is often a RAM issue.

Try new sticks just to see if it solves the problem.
Click to expand...


I agree, as I have seen ram being the issue on my forums across the net.
I bet if you were to ask everyone on this fourm that is having this exact issue of nvlddmkm crashing if they have removed a stick of ram at a time to see if the problem away, a honest and true reply would be "No"

Why?
I guess taking off your side panel and doing something other than type on the keyboard is too hard in some cases. I guess if the problem persists, someone will try it and get a fix.

Then, maybe we won't be the only ones in here suggesting it could be the ram.

It sure doesn't hurt to try.
I've seen everyone talk about removing this driver, and that driver, and such and such, but nobody has yet to mention they removed a stick of ram or two to see if one is bad.
(Even if your memory program told you it was fine)
 
RyanLucier said:
Hey Heather, just curious, did EVGA ever get back to you at all?
Click to expand...

Not even a whisper.

However, it's good to see the XFX guys sticking their heads into the thread.

You know what would be nice? A comment from someone from Nvidia. :D
 
A comment from nvidia? I think they're busy having a snowball fight in hell right now, might be a while. :)

To be honest though, my reason for making this thread has evolved, to the point where now my goal is to keep indexing and compiling information given by people and eventually bring it to the attention of the guys at hardocp.com, or maybe even other websites. I think if this thread eventually hits 15-20 pages and still no solutions are found, it's going to be real hard for NVidia to ignore the issue, when HardOCP and other sites are starting to write articles about the problem and say "hey, this is for real, you can't ignore it any more"

Plus, considering that the GTX 260/280 is right around the corner, NVidia might be even more eager to address the problem as they don't want unhappy customers to be eyeing ATI's competing solutions...
 
l337*g0at said:
Why?
I guess taking off your side panel and doing something other than type on the keyboard is too hard in some cases. I guess if the problem persists, someone will try it and get a fix.
Click to expand...

Least useful response so far.

In fact, SOME people may be having memory issues, but if, according to your 'wisdom', the cure lies in yanking out memory, then please tell us why the software/driver fixes have worked?

Did the driver magically cure the ram?

I'm willing to guess that the problem MAINLY lies in the drivers, as many people have been discovering.

Among the fixes I've seen:

PSU too weak
Bad stick of ram
Weak rail on a PSU

And the most common resolved fixes:

the driver itself
the driver conflicting with an audio driver
a corrupt installation from a driver update
 
Also, the fact that I've NEVER had a NVLDDMKM.SYS error in any DX9 games while in Vista screams to me it's simply not the memory. Memory doesn't magically target DX10 games only to crash, it's got to be a driver issue, especially since some of the DX9 games I play use 2GB of RAM (and sometimes more in the case of Supreme Commander). It's not like the DX10 games are "stressing" the memory harder.. Memory is memory, and DX9 and DX10 games utilize it equally.

I *never* have had a crash with a DX9 game in Vista the past month since I came back to vista to try it out now that SP1 has matured, so I just can't help but say Gee, the crashes only come in DX10... Might be related to DX10 or some software related to DX10. But hardware? doubtful...
 
I would be interested in trying to duplicate your concern, if I can do that then I could submit the issue to NVIDIA. I should have most of the components in your system (minus the sound card and maybe a few others). Please PM me the details of your set up (including windows version, BIOS versions for most components) and I will attempt to reproduce your situation here internally.
 
RyanLucier said:
Also, the fact that I've NEVER had a NVLDDMKM.SYS error in any DX9 games while in Vista screams to me it's simply not the memory. Memory doesn't magically target DX10 games only to crash, it's got to be a driver issue, especially since some of the DX9 games I play use 2GB of RAM (and sometimes more in the case of Supreme Commander). It's not like the DX10 games are "stressing" the memory harder.. Memory is memory, and DX9 and DX10 games utilize it equally.

I *never* have had a crash with a DX9 game in Vista the past month since I came back to vista to try it out now that SP1 has matured, so I just can't help but say Gee, the crashes only come in DX10... Might be related to DX10 or some software related to DX10. But hardware? doubtful...
Click to expand...

I can confirm this is NOT just a dx10 problem. I have had this issue playing games that are dx9 as: GTA:SA, GTA:VC, CS:S, TF2, NFS:C as well as dx10 games.

This issue is NOT just with Vista either. It seems to happen more on Vista, but it has been seen on XP machines too.

Other places/forums have more information you are lacking.
I suggest trying everything possible even if it does not seem legit to you other than putting the problem on something you nor anyone on the forums can fix.
 
Hey XFX, here's the specs of my system you asked for. Pm'd you as well. If I'm missing anything you need, let me know.

- EVGA 680i A1 motherboard, bios P31 (onboard sound is disabled)
- Q6600 @ 3.2ghz (400fsb x 8), 1.44v - no difference if it's stock though
- 2x1gb OCZ SLI memory, PC6400 DDR2 @ stock 4-4-4-15-2T timings, 2.1v
- 2x500GB western digital WD5000AAKS hard drives running in RAID-0
- DVD burner
- Silverstone Zeus 850W power supply
- Creative X-Fi Elite Pro
- 2xEVGA 8800gtx @ stock 576/900/1350 shader - not OC'd, bios version 60.80.13.00.13
- Dell 3007WFP-HC primary monitor - connected to 1st card
- Dell 3007WFP secondary monitor - connected to 1st card
- Sharp 42" 1080p LC42D62U television connected to the 2nd card via HDMI
- Samsung SCX-4200 laser printer
- Canon i9900 printer
- Microsoft USB internet keyboard pro
- Microsoft Xbox 360 wireless receiver
- Microosoft Xbox 360 controller
- Logitech G7 cordless mouse
- Floppy drive
- Swiftech Apollo GT waterblock with triple 120mm fan radiator externally mounted to the case.
- Gigabit ethernet wired to D-Link DIR-655 router, firmware 1.11
- APC 1300VA UPS power backup

Vista SP1 and all updates available on windows update installed.
 
Apallohadas said:
What Power supply?
Click to expand...

Corsair TX 750W ATX 12V psu, should be plenty.

However, after much investigation with my friend, we may have found a solution. Though I don't know the details since it was solved while I was not around. Here's what I know: he installed Vista and got the major updates for it. Had all these problems. Looked into numerous solutions... got me to post here for him again :cool: then realized he was sitting on a bunch of updates for Vista and he just installed them all... Then the problem stopped (at least during the past few hours).

I'm sorry, I don't know which windows patch it was. He was sort of vague on the whole solution since he's been playing games nonstop for a couple hours and I think he just installed them all.

EDIT: The error is still present when playing Age of Conan specifically. :eek: I'm not sure if he's tried other DX10 games yet. AoC boards are saying pretty much to use only one card still. :rolleyes:
 
I have the error & crash on DX9 & DX10 games. Not much fixes it... only periods of time where I can play peacefully.
 
XFXSupport said:
I would be interested in trying to duplicate your concern, if I can do that then I could submit the issue to NVIDIA. I should have most of the components in your system (minus the sound card and maybe a few others). Please PM me the details of your set up (including windows version, BIOS versions for most components) and I will attempt to reproduce your situation here internally.
Click to expand...

That would be immensely helpful, guys. You can also loop me in at [email protected]. I have a customer right now who is having the problem, and can give his specs as well.
 
one thing I've noticed is information pop-ups seem to be the culprit of some of the crashing. Playing world in conflict, at the main menu screen, if I hit the volume up or down button on my keyboard, it results in the driver stop responding and restarting - this is also precisely when a popup occurs on my screen, telling me the volume would be raised (I don't actually see the popup, but it WOULD have shown up if it didn't crash.). It doesn't hard crash in a BSOD though.

Something to ponder, especially since the crash hasn't happened in DX9 games when I raise and lower the volume to the best of my knowledge.
 
This is what I've come up with so far, and any of you can add your observations or any information you have to what I've got.

* This problem happens in both Vista and XP, but only because XP seems to be more fault tolerant.
* The problem seems to be linked, at least in part, to TDR (timeout detection and recovery) features in Windows. When TDR works, the Microsoft Windows Display Driver Model (WDDM) is reinitialized and the GPU is reset withouth causing a hang or crash. When TDR fails, the program will crash to a black screen, the system will blue screen, or the entire system will become non-responsive.
* Correcting the problem seems to entail ensuring that DMA buffer completion takes no more than 2 seconds. Any longer starts the TDR process. This is not something we can fix here, but has to be done from the card manufacturer's end.
* According to Microsoft, if DX9 or 10 runs at a low FPS, it increases the chance of GPU reset. Flushes can be inserted in the graphics code to alleviate the problem, but this must come from the software company itself.
* ATI has its own version of the problem with the atikmdag.dll error.
 
Solve my problem with 8800 gtx's. More than likely heat is the issue because as soon as I clock my 8800 gtx's from 630 overclock factory to 600 it was fine...no more driver crashes. Kinda stupid when factory overclock card doesn't do what it was factory overclocked to.
 
I had that BSOD happen twice right after I built my new rig and was testing it with the latest drivers. Someone on the reviews on NewEgg mentioned that that EVGA 8800GTS KO (factory overclocked) would get too hot and that was the cause of the problem. So I downloaded the EVGA Precision tool and had it set the GPU fan to 75% or above. After I did this I have not seen a blue screen or crash since (about 3 weeks). I'm convinced its a heat issue, or at least that the problem is worsened when the card is stressed.
 
I've had my EVGA 8800GTX KO replaced by RMA due to reheating and have set the fan at 90 to 100% on both cards. I think I'll go about underclocking for now until I get one of the next-gen GPUs.
 
cybereality said:
I had that BSOD happen twice right after I built my new rig and was testing it with the latest drivers. Someone on the reviews on NewEgg mentioned that that EVGA 8800GTS KO (factory overclocked) would get too hot and that was the cause of the problem. So I downloaded the EVGA Precision tool and had it set the GPU fan to 75% or above. After I did this I have not seen a blue screen or crash since (about 3 weeks). I'm convinced its a heat issue, or at least that the problem is worsened when the card is stressed.
Click to expand...

I think you're close.

Once the heat cranks up the FPS drops to the point there the driver failure happens.

So the heat is leading you to a symptom of the ailment, that being the driver failure and recovery.
 
cybereality said:
After I did this I have not seen a blue screen or crash since (about 3 weeks). I'm convinced its a heat issue, or at least that the problem is worsened when the card is stressed.
Click to expand...

I, too, think it might be a stressing issue. I'm working actively with XFX right now in an attempt to come up with a solution (eVGA is still not responding).
 
Folks,

I HAD this problem for a while. I tried all sorts of software updates and rollbacks. Nothing worked. The nvlddmkm error would pop up with some regularity. In fact, I could ensure it would happen with "Company of Heroes, Opposing Fronts". (I'll post my specs later). My quad core had been at 3.2 (a q6600, stock at 2.4 ghz). It had tested 24 hours as stable. Once I dropped back to stock, the nvlddmkm errors were greatly reduced, but would still occur. Finally, I added ANOTHER fan.

I built a special bracket for the 120mm fan; it aimed the airflow directly at the heatsink on the motherboard over the northbridge, with some also angled just under the cpu socket.

I have not had an nvlddmkm error since. I am back up to an overclock of 3.0. (Just too lazy to stop the boot at the BIOS to run it back to 3.2.)

Did I mention no nvlddmkm error since?

This has led me to the conclusion that it was a heat issue in my specific case. Now, if you're really interested and not just a poser, I'll drop the other shoe: my system is very cool. The fan I added is the SEVENTH fan in my case. My CPU temps are 47C in coretemp 0.98.1. My graphics core is 55C using Rivatuner 2.09. I ran prime95, 3dMark06, Pi, etc., all showed a rock-solid system. Adding the extra fan was not indicated by any diagnostic. I did so because the way my components were located seemed to create a minor dead air area between the video card heatsink and the cpu heatsink. Right where the two met there was a dead zone. Also, the 8Gb of memory I knew would create more load on the northbridge.

Take all that for what it's worth. Here are my specs:

Vista Ultimate 64, sp1
Antec p182 case
Seasonic M12 600 Watt power supply
Abit IP 35 pro motherboard
Intel q6600 g0 cpu
8Gb (4x 2Gb) G Skill 6400 ram
evga 8800gtx gpu
2x 500Gb 7200.10 barracuda harddrives
Plextor Px-810SA DVD drive
Samsung DVD drive (forget the model number)
A 50in 1 USB card reader.

For cooling I have the above-mentioned 7 fans. Various Noctua and Nexus (Nexi?). Two dedicated inflow fans in the front of the case and 2 dedicated exhaust fans. I've taped shut all other cracks, openings and case orifices. I'm also using a thermalright Ultima 90 on the cpu with a 120mm fan, on the 8800gtx is a thermalright hr-03plus with a 92mm fan running on a profile on Rivatuner. I've tweaked all the fan speeds via the BIOS. My machine is VERY quiet. (Check on silentpcreview. My approach was a lot of airflow at low fan speeds.) I had already performed the "bolt mod" to the motherboard. Arctic Silver 5 and Arctic Ceramique used variously throughout the machine.

Take this information and use it you'd like.

Regards,
Ken
 
Status
Not open for further replies.
Back
Top