nvlddmkm stopped responding ?

[thilanliyan]

Originally posted by: WT
Already rolled back to 163.86 but couldn't find any older archived driver on the Nvidia website. Might as well Google it and hope for the best.

Here's the 163.44 series (choose your OS):
http://www.gamershell.com/search/?q=163.44

Don't know if it'll work for you but give it a shot. And remember to update DirectX as well.

 

[ZK2007]

Good luck on your endevors, WT.. I tried changing drivers back and forth, whql, beta and even older drivers, Formatting and starting with latest whql, taking out sound card, changing vidcard to another slot, going to XP(instant BSOD) and as i said.. all failures. 20 crashes a day would have been better on me than what I was getting.. 1 a second until BSOD in Vista. IP35 Pro here for 3 weeks including a new cpu upgrade.. otherwise same ram and video card.. never a crash. Using the same drivers that gave me hell on my Nf590 board. I've been playing HL, HL2 EP1 and 2, Portal, TF2 all for as long as i want, when I want and system is stable. Honestly I'd like to see how many Intel chipset users have these nvlddmkm crashes compared to those using nvidia chipset boards. I'm playing games now that, with the NF590 board, I couldn't even get to run a few seconds in to the game even if I managed to not crash before it got started.

 

[WT]

FIXED !!!! Yanked out the Buffalo Firestix and popped in a pair of Super Talent DDR2-800's and all is well again. I've played two straight days now so it looks like the Buffalo RAM that I paid so dearly for is my faulty hardware. Thank God that horrible experience is now over with. Good luck to anyone else with this issue, and if you still see this error message on your PC, try and run a Memtest on your RAM.

 

[Maximilian]

This is f**king bulls**t!!! Ive been all over the internet looking for a solution to this problem, all i see is crazy solutions that dont work or people blaming hardware, if it was hardware then there wouldnt be so many people with the same goddamn problem, and it likely wouldnt "recover".

I have this problem with company of heros, i installed the latest drivers, and still have the problem BUT it sorts itsself faster now, usually before the enemy gets the chance to kick me, but not always. This is trash, i cant believe people are still having this problem.

Is there a solid solution to this?? I have vista x64 and an 8600m GT with the 165.01 drivers.

 

[thilanliyan]

I don't know about a solid solution...all you can do right now is to try whatever you can within reason. For me it was a software fault, for WT is was hardware...who knows. I think it's mostly a software issue but...who knows!!

 

[Noirriver]

Hi guys,

Basically after not being able to run Planet Extreme for more than a few minutes or World in conflict, I've been playing so far for 3 days with no issues (fingers cross)

I read everything from driver, DX10, temperature and voltage related, etc?and for me nothing was working

But after tinkering with this for a few days ?(I think) the issue appears to be related (at least for me) with the Nvidia driver setting for "Textures" and "Adjust image setting with preview".

Running with 163.69, without even reinstalling drivers etc?
I have:
?Adjust image setting with preview? : Let 3D application decide (yes, for me some times even just asking the driver to show the preview would make it crash!!!)
?Manage 3D settings? (Global settings): main thing here is around ?Texture? management.
- Texture filtering (TxF) ? Anisotropic: Off
- TxF ? Negative LOD Bias: Clamp
- TxF ? Quality: Performance
- TxF ? Trilinear optimization : Off
-
You could play with these setting more? I was just happy that I was able to play without crashing so I haven?t gone for the ?fine tuning? yet ?

But (at least for me) it appears to be ?Texture? management related...

Bets of luck!

E6850, 4GB ram, Vista Ultimate, 8800 GTX XFX

 

[Maximilian]

Originally posted by: Noirriver
Hi guys,

Basically after not being able to run Planet Extreme for more than a few minutes or World in conflict, I've been playing so far for 3 days with no issues (fingers cross)

I read everything from driver, DX10, temperature and voltage related, etc?and for me nothing was working

But after tinkering with this for a few days ?(I think) the issue appears to be related (at least for me) with the Nvidia driver setting for "Textures" and "Adjust image setting with preview".

Running with 163.69, without even reinstalling drivers etc?
I have:
?Adjust image setting with preview? : Let 3D application decide (yes, for me some times even just asking the driver to show the preview would make it crash!!!)
?Manage 3D settings? (Global settings): main thing here is around ?Texture? management.
- Texture filtering (TxF) ? Anisotropic: Off
- TxF ? Negative LOD Bias: Clamp
- TxF ? Quality: Performance
- TxF ? Trilinear optimization : Off
-
You could play with these setting more? I was just happy that I was able to play without crashing so I haven?t gone for the ?fine tuning? yet ?

But (at least for me) it appears to be ?Texture? management related...

Bets of luck!

E6850, 4GB ram, Vista Ultimate, 8800 GTX XFX

I will try these settings with my 8600m GT and 165.01 drivers tomorrow and report back if i still get a TDR error.

 

[PepperBreath]

In case it hasn't been mentioned yet...this is what I did and it solved the problem for me.

System was running fine until one point I couldn't play more than a few minutes of STALKER, Oblivion, or anything else without the screen freezing, going black, then "waking up" and running again only to repeat the process. When I'd exit the game, I'd have a little exclamation point on the taskbar saying that that nvlddmkm had a problem and was shut down but was able to be started again (obviously why my games would start working for another minute or two).

Anyway, after doing some searching, I tried one method that worked...for a short time. I "tweaked" the procedure a bit and my PC has been working perfectly ever since.

This is what you do.

1. Grab the newest drivers from Nvidia's site beta or not, doesn't matter. Whichever release you want installed.
2. Download Driver Cleaner Pro and install it. Don't run it yet.
3. Uninstall your current drivers. On my system, the screen went blank and never returned. If this happens, wait about 10 minutes and then hit "Return" on your keyboard. Why? Because at the end of each driver uninstall, they always present a dialog to reboot the system. Your system continued to uninstall the drivers even if you were not able to see it do so on the screen.
4. Tap F8 repeatedly on the keyboard immediately after the POST screen but before the Vista screen. If you were on time, you'll get the Vista bootup menu. Select Safe Mode (first option).
5. A window should pop up asking for drivers, select "Ask me again later". Start Driver Cleaner Pro. On my system, it never completes. Just let it delete what it can and if it stops. Just close it and reboot.
6. Boot up into Vista (non-safe mode). Again, a window should ask if you want to install or search for drivers, select "Ask me later". Install the drivers you previously downloaded.
7. Test it out. Since I've done this, I haven't had a single problem since.

Every system is different obviously so I can't guarantee this method (I'm not nvidia so I don't have to. ). However, I'm satisfied and I've already played and completed a few games since doing this without a single problem. Good luck.

 

[Noirriver]

Hi guys,
I?m afraid to inform that regardless of my previous Post that early last week I started to get those errors again. I did a bit more poking and now (take it with a pinch of salt) I think the issue is between the Freq. of the processor, the RAM frequency and the GPU freq... not temperature, nor software (could be BIOS though)

Now I have 2 setting that under some game stress testing (ie WIC for 8hrs+, Lost planet 6hr+, etc) seem to work:

1. I have 4GB of 800 RAM, I have under-clock it to 667 and now I can set back all my driver setting to high and no crashing.

or

2. leave RAM at 800 but ensure processor (I have E6850) is at 1600 instead of 1333.

What it seem is that for "high" end machines with some NVida chipsets some DX10 applications have a hard time when the RAM - Processor freq. relation is not exactly 2:1 (ie RAM 667 - Processor 1333 or 800 RAM - Processor 1600)

Is simple to try any way and (so far) no crashing for me for ~8 gaming days in a row...I?m back to enjoying games!

 

[chizow]

Originally posted by: Soviet
This is f**king bulls**t!!! Ive been all over the internet looking for a solution to this problem, all i see is crazy solutions that dont work or people blaming hardware, if it was hardware then there wouldnt be so many people with the same goddamn problem, and it likely wouldnt "recover".

I have this problem with company of heros, i installed the latest drivers, and still have the problem BUT it sorts itsself faster now, usually before the enemy gets the chance to kick me, but not always. This is trash, i cant believe people are still having this problem.

Is there a solid solution to this?? I have vista x64 and an 8600m GT with the 165.01 drivers.

I totally understand your frustrations as this was something that I dealt with on and off with my system since I built it. In every case, the root problem turned out to be faulty memory. That's not to say other factors can't cause problems, as I'm pretty sure NV chipsets and Vista not getting along well can cause problems also.

Some things you can do to try and cut down on TDR errors:

1). Update the BIOS. Check frequently for any updates, especially if you're running 2GB+ or all 4 DIMMs populated.

2). Get all the relevant Hot Fixes for Vista. There's at least 3-4 Hot Fixes that deal with TDR/Memory Management/Games/4GB in Vista. Some update through Windows Update, others won't be default updates until SP1.

3). Update Drivers to 163+. Even though I hadn't had any TDRs on my GTS (aside from bad RAM on 2-3 occasions), Vista's desktop loaded "strangely". Hard to describe, like it had black bars and such before the image fully appeared. After 163s that went away.

4). Update chipset drivers, relax RAM speed and timings. Especially if you use NV chipsets. Really regret going with nForce chipset as its been the cause of a lot of problems for me (X-Fi, dead RAM etc).

Hope that helps.....

 

[Mem]

Another point I like to make,Nvidia firewall can cause issues too in both XP and Vista so would advise not to install that (where available on nVidia based boards).

The only game so far that has given me any real problems was The Witcher( kept crashing ,running in Windowed mode fixed the problem for me,common problem if you check their forums).

Anyway tried about 40+ games so far in Vista x64 all working great.
Its very easy to blame OS or even drivers but sometimes harder to troubleshoot and find the real cause of the problem,at least I can blame Atari for my problem with The Witcher.

 

[Maximilian]

After having sat and played CoH all day i dont remember having any TDR errors after making the suggested changes, im tired to the point of stupidity.... Oh right ok scratch that, i just checked the task thing at the bottom right and that little yellow warning triangle is there, so i guess i had a TDR error at some point! Balls.... oh well.

Good thing is that the game runs fine after the error occurs and recovers quickly now with newer drivers, basically the same situation as before, nothing has changed.

Bad thing is Company of heros crashes sometimes, when i try and join a certain game it will crash, when i exit to windows it sometimes crashes. I dont know whether or not to attribute this to relic and it being buggy at the moment or if its the drivers causing this. This is also the same as before, just didnt think to mention it before.

 

[taltamir]

The driver crashing and recovering is a SYMPTOM not a CAUSE. Before nVidia (less then a year ago) and AMD (about a year ago) implemented recovery it would have simply resulted in a memory dump / complete freeze of the system. Now it restarts the video card only. Which is GREAT!

The reason you see so many solutions with some working for some people, some for others, and some having no solution is because we are describing a symptom, the absolute most common symptom, the "my video card crashed (and recovered)" symptom.

You just need to try the various tests and see what happens. You could have faulty ram, faulty video card, it could be an issue in vista, it could be an issue in drivers, it could be an issue in the game itself (I know of a few games that had such issues and fixed them, or some that didn't)!

Basically if you are getting it your video card crashes... here is what you should do:
1. Run memtest, if there is even a single error your ram / motherboard is defective. Replace ram.
2. check temperatures and airflow for video card, if it is burning up then you have your culprit, if you have bad cooling you can down clock your card a little to solve it.
3. Is it only happening on one or two games? Try some other intensive games, if they never crash then it is the game (or the game/hardware combo) and there is nothing you can do.
4. Is your power supply good enough? I built a system last year with a 7900GS OC with athlon X2 3800+ and a 400 watt power supply and it crashed about once an hour, after intensive gaming.. i upgraded to a 500watt psu and it stopped happening. (took me months to figure out the exact problem, i replaced half the computer too!)
5. Your card might be defective. if its not 1-4 then it probably is, RMA it.

After doing all this you should no longer have the problem.

 

[drednox]

Originally posted by: Mem
Also overclocked video cards can cause this problem too,underclocking fixes it.

that is a false statement. it is directly inacurate and/or at best based on heresay and no facts whatsoever and it obviously comes fro ma person taht is compeltely clueless with regards to this issue.


i also had this problem, the following is how it all went down:

in february / march i built a whole new system: core 2 duo 6800, 4 ggis of corsair 1066 twin ram, evga nf680i SLI board, a single evga GF8800 ultra /w 768 megs. s/b x-fi, nice HD and DVD-RW drive to go with it.

everything was cool until about 2 montsh ago when all of a sudden i got that error and all my 3d apps started crashing, even screensavers and websites that were using 3d modeling.

i wasted at 1st about a week and maybe 60 hours updating all drivers, bios-es, manually deletign nvddmkm files, had to learn vista security, account menagement and all taht kidna stuff, cause otehrwise you cant touch all of them anyways. did virus scans, spyware scans, purchased 3rd party software to perform additional scans, etc etc etc. also, beware, the generic "update your drivers" advice in this case is the worst and most wrong thing to follow. updating drivers just makes this worse and happen more often, not less.

after that i finally caved in and backed up any important stuff i had and did a full complete format on my HD. i figured whatever was wrong with it, it all started with a driver update taht i performed when i purchased and installed bioshock. so i thought it was software and thought a clean wipe would fix it.

after compelte hard format reinstall of vista 64 lo and behold it came right back up. on a clean install, nothing but the vid card plugged in and keyboard and using the drivers latest WHQL drivers posted / linked by microsoft to be the latest certified nvidia drivers.

even reformat didnt help.

nvidia support is compeltely useless and rude about it, they flat out deny the issue exists and keep pointing fingers at the rest of the system and microsoft. microsoft keeps pointing fingers at nvidia. and evga, the manufacturer keeps pointing fingers at the both of them, as a result i am still fu@ked along with everyone else havign thsi problem.

after a while i caved in again and blew about $250 and picked up ATI Radeon 2600 XT at fry's.

the nvlddmkm went away but ....

now i have the missing texture problem that is so prominent on ATI cards, especially in dx 10 games. i instaleld the missign texture fix from the ATI website but it seems after the fix you can only run those games in low settings and low rez. i have a widescreen high rezz HD monitor and like to run my games at max. so their fix is pointless. ATI support, like nvidia is apathetic, in denial "its noit us its you" standard PR bullshiht mode.


as a result of all of this, i have a nice $700+ P.O.S. nvidia garbage lying in my drawer and another cheapo card in my comp that runs only half the games i wanna run.

i deeply regret ever beign tempted with switchign to win vista, dx10, and 4 ggis of ram so i had to go 64 bit further minimizing hardware choices and software options as well.

i am sorry this happened to you, my nvidia card seems to be now permamently damaged hardware wise. it now has a lot of weird artifacts, pixelization errors, missing / corrupted textures etc., even durign the short periods of time before the error comes back, nothing is playable on it.

i have no solution for you. i know few people had some limited sucess with the following: geting better quality power supplies and making sure they got high quality current coming across both the 12v rails taht plug into the vid card, rolling back teh drivers a few versions, 1.58 i think or lower + manually deleting ALL the old nvlddmkmk files and some related including from the vista backup and restore dirs, which will require a large amount of wasted time on learnign the ins and outs of vista security and features to accomplish. you have to do this before you install that 1.5xxx driver.

again, that is limited on case-by-case sucess, some have been sucessfull some havent, about 50/50.


good luck man.

 

[taltamir]

Originally posted by: drednox

Originally posted by: Mem
Also overclocked video cards can cause this problem too,underclocking fixes it.

that is a false statement. it is directly inacurate and/or at best based on heresay and no facts whatsoever and it obviously comes fro ma person taht is compeltely clueless with regards to this issue.

[summerized] my video card broke and I tried every software solution and it didn't fix it so I bought a new unbroken ATI card and it "fixed" it[/summerize]

that was horrible drednox... you disputed a perfeclty valid claim, which is "overclocked video cards might crash often" and to back it up explained how your card worked for a while, then it just "magically broke"... proceeded by how no software gimmic you did fixed it, and how nvidia somehow sucks because of it and buying an ATI video card was the fix...

Your fix would have also been to RMA the broken card for warranty and get a non broken one... it has nothing to with any of the things you said... it wasn't the drivers, it wasn't the vista, it wasn't anything.. your card broke, IT HAPPENS!

This error message means one thing and one thing only, for some reason or another your video card crashed... it can be overclocking, a bug in software, a bug in drivers, bad ram, or your video card itself broke... Blaming specific things or treating it like a "known problem that they don't fix" is ridiculous.

 

[drednox]

Originally posted by: taltamir

Originally posted by: drednox

Originally posted by: Mem
Also overclocked video cards can cause this problem too,underclocking fixes it.

that is a false statement. it is directly inacurate and/or at best based on heresay and no facts whatsoever and it obviously comes fro ma person taht is compeltely clueless with regards to this issue.

[summerized] my video card broke and I tried every software solution and it didn't fix it so I bought a new unbroken ATI card and it "fixed" it[/summerize]

that was horrible drednox... you disputed a perfeclty valid claim, which is "overclocked video cards might crash often" and to back it up explained how your card worked for a while, then it just "magically broke"... proceeded by how no software gimmic you did fixed it, and how nvidia somehow sucks because of it and buying an ATI video card was the fix...

Your fix would have also been to RMA the broken card for warranty and get a non broken one... it has nothing to with any of the things you said... it wasn't the drivers, it wasn't the vista, it wasn't anything.. your card broke, IT HAPPENS!

This error message means one thing and one thing only, for some reason or another your video card crashed... it can be overclocking, a bug in software, a bug in drivers, bad ram, or your video card itself broke... Blaming specific things or treating it like a "known problem that they don't fix" is ridiculous.

i do not overclock. i run all my stuff stock. also had plenty of cooling. and i should also add i have a good P/S. also, if you had read, i did post initial driver hang happened when i purchased and instaleld bioshock and it automatically updated my driver. and i also added taht with each driver update these thinsg just got progressively worse, instead of better. i know i make many typos, some intentionally some because i am tired and some because i use them as shorthand and i tend to ignore grammar, but even then, learn to read. go back to school and take some contemporary composition classes or alike. and even moreso, pay attention to what your eyes wonder over.

RMA isnt an option. whos goign to pay for the time my system is down and the drama that will cause while waiting on the stupid RMA ?

RMA will become an option when people that create these fuhcked up products accept some responsibility for them and put their money where their PR and advertising claims are and cover all the damages resulting from their faulty products.

i dont have the luxury of waiting on an RMA. i need my system because i use it for things otehr then gaming as hard as that may be to understand for you and many others.

i dont give a crap if its microsoft, nvidia, evga, as far as i am concerned what would make me happier now is if the 3 of them go and shoot / hang / disembowel / rape themselves and tehir children IRL for making this shiht then denyign that the problems even exist, much less offer any actually usefull technical support.

and its peopel like you that seem to agree with them and buy into it all and make things worse instead of better.

 

[Mem]

Originally posted by: Mem
Also overclocked video cards can cause this problem too,underclocking fixes it.

Originally posted by: drednox

that is a false statement. it is directly inacurate and/or at best based on heresay and no facts whatsoever and it obviously comes fro ma person taht is compeltely clueless with regards to this issue.

Read my earlier posts in this thread you'll see underclocking my Video card by 10 mHz(factory oveclocked card btw) resulted in getting 100% stability in virtually every game(this is for XP and Vista) I no longer get the "nvlddmkm stopped responding message".

I not only backed up my claim but have also proved it in my case,remember there are many causes why games crash,however in my case the one I meantioned was the problem.

Next time you try to insult me have some real facts to prove me wrong,oh wait you don't have my video card so can't prove it,but I can ,I rest my case.Read the next post too.

 

[Mem]

Originally posted by: wazzi

Originally posted by: Mem
Also overclocked video cards can cause this problem too,underclocking fixes it.

in my desprate attempt to keep win vista i took mems advice n down clocked my xfx 8800 gts 320 mb xxx(factory overclocked my A*S) from 580 mhz to 560 n every thing seems to be working fine till now, but this is shit man ,factory overclocked r unstable why the f**k should we buy factory overclocked gpus then n pay extra for them

Yes underclocking video cards (especially factory overclocked)works for some people.

 

[ctark]

If you have an Nforce board get rid of it. I switched to an intel chipset and now my card works fine. It gave me the driver crashing error on 2 video cards.. both ATI and Nvidia.

 

[taltamir]

Originally posted by: ctark
If you have an Nforce board get rid of it. I switched to an intel chipset and now my card works fine. It gave me the driver crashing error on 2 video cards.. both ATI and Nvidia.

seeing as how no nforce board ever gave me that problem I am going to just chulk it up to you having a BROKEN BOARD.

Did you try replacing it under warranty or did you just decide "nvidia suxor" and buy an intel board instead?

Also, who was the manufacturer? was it someone known to make unreliable boards like ECS or Abit (I know abit used to be considered the best, but now they are just unreliable)?
Or was it someone reliable like gigabyte, asus, or msi?

 

[chizow]

Originally posted by: taltamir

Originally posted by: ctark
If you have an Nforce board get rid of it. I switched to an intel chipset and now my card works fine. It gave me the driver crashing error on 2 video cards.. both ATI and Nvidia.

seeing as how no nforce board ever gave me that problem I am going to just chulk it up to you having a BROKEN BOARD.

Did you try replacing it under warranty or did you just decide "nvidia suxor" and buy an intel board instead?

Also, who was the manufacturer? was it someone known to make unreliable boards like ECS or Abit (I know abit used to be considered the best, but now they are just unreliable)?
Or was it someone reliable like gigabyte, asus, or msi?

There's countless threads including an acknowledgement from Creative about problems with nForce boards and the X-Fi. I've experienced it myself and its resulted in having to pull my X-Fi indefinitely. Something going on with the chipset, Vista and resource allocation (IRQs, memory addresses etc).

I've also had one nForce board (Asus) kill 6 different sticks of RAM; swapped out the board and everything ran fine for a few months but started having problems again so I swapped out to 2 x 2 GB. Either the chipset itself has issues running 4 dimms (I had to flash the BIOS before 4 dimms would even load Windows) or board makers are cutting corners and using lower quality components for the 2nd pair of dimm slots or there's a problem with power regulation/providing enough voltage. Really not sure but I do know buying an nForce board instead of an Intel chipset is the only thing I regret about my latest build.

 

[taltamir]

Um, creative is THE PLAUGE!

Creative has the absolute worst drivers EVER made for anything. Creative simply does not CARE. From 2002 to 2005 it was impossible for me to install their latest drivers DIRECTLY without first installing older drivers from the original CD because the drivers gave me the following message:
"The license for this software is for upgrades only, please install the original version that came with your software CD"... then they made it possible to install the DRIVERS, but not their control panels and software.. because everyone just wants to steal their drivers to use with other sounds cards... Their auto update software was buggy and got corrupted often where only a reformat will allow it to work again. The drivers had a ton of problems installing...and I am not just imagining it, there are threads on threads where people have all come to an agreement that creative just has the worst drivers ever made.

Not to mention it destroying windows, and breaking everything else...

If a creative card doesn't work its solely the fault of creative, not nvidia!



And you should have replaced that asus board after the second stick broke... it isn't the chipset's fault, its a faulty board the provides excessive amounts of voltage...

I had once gotten a board that killed 3 hard drives by sending too much voltage on a SATA port... I replaced the board and it stopped breaking my drives. Broken hardware happens!

 

[Garfield3d]

Add me to the list of discontented. Here are my system specs:

Intel Q6600 (G0 stepping, pack date 10/14)
Zalman 9700
Gigabyte P35-DS3R (rev 1.0, F6 BIOS)
Corsair Ballistix PC-6400 (5-5-5-18)
Western Digital 500 GB AAKS
Geforce 8800 GTS (320MB)
Windows Vista 32-bit
Antec Earthwatts 500W (manufactured by SeaSonic)

My computer is almost 5 days old now. I'm receiving the "nvlddmkm has stopped responding and has successfully recovered" with only 3d-games; I've never had the error while at desktop (using Aero) or in 2D mode. When I get the error, the screen freezes for maybe 2-10 seconds, then it will black out. The results from there vary: If I'm playing something like World of Warcraft, the screen will just flicker and come back without a hitch. For WoW, it really looks like that the screen is flickering like it's changing resolutions (resolution changing and crashing both reset the driver, I imagine). In Crysis or Lost Planet, though, the game usually blacks out without recovering and from there I either need to Alt+F4, alt-tab and kill the game manually, or wait a minute and hopefully get an error message. In all situations, the system tray gives me that innocuous "nvlddmkm has stopped responding and has successfully recovered" message.

In my case, I don't think this is a memory problem. To test my memory, I overclocked the memory from 800 mhz to 1066 mhz then ran Prime95 (version 25.4, torture RAM test, round off checking) overnight (a span of 6 hours). Afterwards, I ran the Windows Memory Diagnostic tool. Neither case found any errors. I then downclocked the RAM back to 800 mhz. However, the error continued to persist in all 3 games (WoW, Crysis, and Lost Planet).

The CPU doesn't seem suspect either. I haven't overclocked my Q6600 (G0 revision) and a max FPU torture test found no errors as well. I don't know how to test the motherboard itself, but I imagine that the Prime 95 test must be doing its best to murder the Northbridge.

I don't think heat is a problem either. Depending on the ambient temperature (anywhere from 74-82 F), my idle is anywhere from 36-40 C. On load, my temperatures jump up to about 59-62 C. I've tried to run the games with the Zalman 9700's fan at its highest setting as well as taking the side of the case off (that usually lowers the idle CPU down to 32-34 C). However, the errors will still show themselves, sometimes about 5-10 minutes into the game, and sometimes more towards an hour into the game.

The hard drive seems okay too. Running 3-hour extended tests from WD's diagnostic tools didn't bring up any errors. SMART says everything is fine, but I've never trusted SMART to find anything. I also installed Windows Vista on a complete format. Like everything else in the computer, the hard drive is also new.

I highly doubt this is an issue of driver conflicts. As mentioned, this computer is only 5 days old and it's a new install of Windows Vista. I tried to install updates and drivers in an order that would least minimize conflicts: Intel INF Chipset utility first, then Windows critical updates next, LAN driver third, audio/video drivers fourth, and Windows important updates 5th. I also installed the Windows Hotfix for 2GB addressing files.

I'm running the F6 BIOS for my Gigabyte P35-DS3R. I tried the F7d beta BIOS at one point, but it was unstable for me, so I went to F6. I could try to "downgrade" to F4, but such a move sounds more like a blind shot at the dark rather than an educated solution.

I've tried running different driver versions. I started out in the beginning with 163.69. Then I tried 163.75. Then I tried 158.45. 158.45 seemed to last longer before giving me the error. It also performs much slower in DX10.

The only things that I don't think I can fully account for are the motherboard and the power supply. I imagine Prime 95 would spit out errors if there were motherboard problems. As for the power supply, Gigabyte's BIOS doesn't provide specific voltage information. I imagine if the power supply was the problem, I would be experiencing hard crashes that require complete reboots. Or, I would be seeing errors in more components than just nvlddmkm.

I would like to note that the OS hasn't been crashing and no other programs have been crashing. Vista isn't locking up, and as far as I can tell, there isn't anything unusual happening with the system. If it weren't for the nvlddmkm error, I would have said that my computer is doing perfectly fine.

I find it very hard for me to imagine this as anything other than a video card / driver error. For those that are suggesting that this is a memory error, a heat error, an overclocking error, an old-driver-residue error, or whatnot, what sort of other diagnostics do you have to suggest? I've done all I can imagine to check for failures in those other areas, and every benchmark, downclock, or shot at the dark continues to produce nvlddmkm errors. And, everything else and every diagnostic benchmark seems to be doing fine. It's always just nvlddmkm.

--Garfield3d

 

[taltamir]

The "screen flickering like it is changing a resolution" effect you get is when the video card crashes, completely, and instead of getting a BSOD the new nvidia drivers restart the video card only and give you that message. All those tests people mentioned are to see WHY the video card crashes... but the very first thing you should to is replace the video card with an identical model and see if it solves the problem... 95% chance you got a broken video card.

prime 95 tests your CPU not your memory! It also cannot tell you if you have a problem with the video card related parts of the motherboard. It only tests the CPU or the motherboards interaction with the CPU.

Windows memory tester is crap.. Use memtest86+
http://www.memtest.org/

But I doubt you have memory trouble since you would be getting BSOD and random crashes on OS, especially on a serious OC like before..

I would recommend that you try replacing the video card with an identical one, it is the most likely culprit. If not then the motherboard, if not then the PSU... one of them is defective.
100% chance that you have a defective part... it cannot be the drivers! You have a fresh install of vista with current drivers, why would only you get crashing in 3d games due to any reason other then faulty hardware?


Look here for info about this problem:
http://forums.anandtech.com/me...=2116981&enterthread=y

 

[ZK2007]

Alright guys I am here to report back with my info about curing my Nvlddmkm error problem. It's been since late September since I fired up my new system. I've yet to get a BSOD OR Nvlddmkm crash error. I've not even had so much as a lockup of any sort. All the same hardware that didn't work with that Asus P5N32SLI Premium SLI WiFi/AP Solo NF590. I am running now, an Abit IP35 Pro. Same OS, same ram, same 8800GTX video card, same everything but Nvidia chipset board and cpu, went from E6700 to E6850(absolutely no reason to cause those errors). I've played every HalfLife game there is, even beat EP2.. I've beat Portal, played CoD4, Hellgate, Crysis Demo.. 3DMark06(wouldn't even try to start w/nvidia chipset board). Most of my HalfLife games wouldn't get past the Options menu to select Play. With this said, I firmly believe it's the nvidia chipset mb that was the issue. Folks keep blaming the ram and/or video card.. it's not! Folks keep blaming Vista.. Vista is flawless and I love it! I have the stability of XP now and I can finally feel no regret running something now just to see the computer crash out from so many nvlddmkm crashes.

Nvidia chipset board = NVLDDKM crash
Intel chipset board = No NVLDDMKM crash

From what i could see by just scanning sigs.. i didn't see one user with an Intel board complaining about the crashing. Just my 2 cents and my personal experience. I, for one, am extremely happy with my system now that I can run any game I choose or heck.. load any jpg of my choice and not crash the video driver!