Constant nvlddmkm errors

[Axtasium]

I've been fighting it for weeks now and wanted to know if anyone has any ideas for anything I can try or possibly a known good driver from nvidia.

I have tried using DDU and tried a few different drivers and they all had issues. I've tried the 3 most recent drivers and they have all had this issue. Windows preinstalled one that was about a year and a half old and it was having the error every 5 seconds. Not sure what driver to try or if that even is the issue. I have also tried running sfc /scannow and some other checkers (including the built in memory test on windows) and none came back with any issues. Sometimes the issue is non-existent and other times its constant. Everything crashes... games, wallpaper engine, and itll even show up on my desktop from time to time. Any ideas?

I attached a screenshot of wallpaper engine spewing the error and 2 screenshots of event viewer. Stellaris also pops up a log saying something else but I cant be bothered to wait for it to crash just for a picture.

 

Specs:

MotherBoard: Asus Hero VIII (X570)

CPU: Ryzen 9 3900x

GPU: Zotac RTX 3080 10GB

RAM: 32GB 3600mhz

PSU: 850W Gold

[Whatisthis]

Try these steps: https://www.thewindowsclub.com/description-for-event-id-0-from-source-nvlddmkm-cannot-be-found

[gen_angry]

My wife's build had these issues with her 5700G and 1060 build. Constant GPU driver crashes just like those. What fixed it was me setting her RAM speed to 3200, the memory controller was just not stable at 3600 MT/s. Since setting that, it hasn't crashed once in about half a year. She lost the silicon lottery I guess.

 

A 3900X's memory controller is rated for up to 3200MT/s speed, anything higher is considered an 'overclock'. I would give that a try first to see if things improve.

[Axtasium]

8 minutes ago, gen_angry said:

My wife's build had these issues with her 5700G and 1060 build. Constant GPU driver crashes just like those. What fixed it was me setting her RAM speed to 3200, the memory controller was just not stable at 3600 MT/s. Since setting that, it hasn't crashed once in about half a year. She lost the silicon lottery I guess.

 

A 3900X's memory controller is rated for up to 3200MT/s speed, anything higher is considered an 'overclock'. I would give that a try first to see if things improve.

Yeah, guess thatll be the next thing i try then. its just very weird to me as it started recently (about a month ago). been using my memory set to 3600 since i got my computer and never had an issue before. I know its not power either as the issue pops up even if the gpu isnt drawing anything. Gonna try reseating my memory and gpu first just to make sure.

[Somerandomtechyboi]

40 minutes ago, gen_angry said:

My wife's build had these issues with her 5700G and 1060 build. Constant GPU driver crashes just like those. What fixed it was me setting her RAM speed to 3200, the memory controller was just not stable at 3600 MT/s. Since setting that, it hasn't crashed once in about half a year. She lost the silicon lottery I guess.

Likely a ram issue or a mobo issue

Cezzane has a beast of an imc (5000+ ddr4 1:1 fclk) so problably just crappy rams or crappy board

 

For non apu zen2 they tend to max out around 3800 especially on the dual ccd cpus like the 3900x/3950x so lowering ram speed may help

 

if lowering ram speed fixes your issues see if you can run 3600 but you manually set vsoc somewhere around 1.15-1.2v, usually sweetspots around there apparently. You can also try setting vdimm to 1.4v

 

 

Btw can you take a screenshot of cpuz ram tab? Just need to see the rams xmp profile, if it happens to be 16-16-16 primaries you are problably hitting imc limits cause bdie is really stressful on the imc

[gen_angry]

32 minutes ago, Somerandomtechyboi said:

Likely a ram issue or a mobo issue

Cezzane has a beast of an imc (5000+ ddr4 1:1 fclk) so problably just crappy rams or crappy board

 

For non apu zen2 they tend to max out around 3800 especially on the dual ccd cpus like the 3900x/3950x so lowering ram speed may help

 

if lowering ram speed fixes your issues see if you can run 3600 but you manually set vsoc somewhere around 1.15-1.2v, usually sweetspots around there apparently. You can also try setting vdimm to 1.4v

 

 

Btw can you take a screenshot of cpuz ram tab? Just need to see the rams xmp profile, if it happens to be 16-16-16 primaries you are problably hitting imc limits cause bdie is really stressful on the imc

 

She's got an Asus B550-I Gaming board and I believe 32GB G.Skill Ripjaw V 3200s now (I replaced it very recently, she wanted more RAM for Stardew mods, lol). But it was previously 16GB G.Skill Ripjaw V 3600s for the first 7-8 months of her build. I wouldn't doubt it was just something being crap somewhere but the solution did work. Performance 'loss' was not even noticeable.

 

I know Cezanne does generally have a decent IMC (have built a bunch during the GPU shortage with them all posting no problem at max speed of whatever gets thrown in there) but even still - it's only rated by AMD for up to 3200. It is possible to get one from the far end of the bell curve.

 

Regardless, it's a good step for OP to try first. Hopefully that's all it is because other solutions are likely not to be so 'simple'.

[Somerandomtechyboi]

2 minutes ago, gen_angry said:

it's only rated by AMD for up to 3200. It is possible to get one from the far end of the bell curve

Ratings are extremely conservative with the exception of mobos where theyre literally advertising stuff that wont even run on the thing, but fair enough you could have lost the lottery so hard that it wont even go past 3200. I do suggest trying to raise vsoc since cezanne scales alot with vsoc and you can assume vcore = vsoc = max ~1.45v if those igpu oc tutorials are anything to go by

if you are curious about cezanne imc then heres some sources

Spoiler

 

(Down in the comments)

 

 

[Axtasium]

3 hours ago, Somerandomtechyboi said:

Likely a ram issue or a mobo issue

Cezzane has a beast of an imc (5000+ ddr4 1:1 fclk) so problably just crappy rams or crappy board

 

For non apu zen2 they tend to max out around 3800 especially on the dual ccd cpus like the 3900x/3950x so lowering ram speed may help

 

if lowering ram speed fixes your issues see if you can run 3600 but you manually set vsoc somewhere around 1.15-1.2v, usually sweetspots around there apparently. You can also try setting vdimm to 1.4v

 

 

Btw can you take a screenshot of cpuz ram tab? Just need to see the rams xmp profile, if it happens to be 16-16-16 primaries you are problably hitting imc limits cause bdie is really stressful on the imc

its just set to the XMP profile. built this pc when i graduated so it was about 3 years ago? never had any issues before with nvlddmkm errors so. but yeah its not 16-16-16.

[Whatisthis]

1 hour ago, Axtasium said:

its just set to the XMP profile. built this pc when i graduated so it was about 3 years ago? never had any issues before with nvlddmkm errors so. but yeah its not 16-16-16.

But it is 3600mhz (1800x2). Drop it to 3200 and see if the issues persist. 

[Axtasium]

6 minutes ago, Whatisthis said:

But it is 3600mhz (1800x2). Drop it to 3200 and see if the issues persist. 

yeah, it kinda just shows up when it feels like it. went ahead and did a bios update and updated the gpu drivers to the latest so waiting to see if its still there, if it is, im going to reseat my gpu and memory. if its there after that, ill drop it to 3200 and test that.

[Axtasium]

Starting to wonder if maybe something is up when it wakes up from sleeping after being asleep awhile as that seems to be when it has the most issues. Zero issues yesterday after updates. woke it up today, 5 minutes later i get a nvlddmkm error. gonna restart my computer and test that, but not sure why it would do that.

 

Restarted my pc and it seems to be gone again. Not sure why, but maybe something is up with it when it wakes up though i have zero idea what could cause that whether it be windows itself or a driver behaving funny. I also did a quick sleep and wake up and no issues so I don't know the root of the issue.

[Axtasium]

I turned off "turn on fast startup" and I haven't had an issue since. I'm not sure what exactly was going on, but since i turned that off and restarted my pc, there have been zero issues. seems like its some wack windows issue atm. Ill update this post if something does decide to flip out again but havnt had a single issue in the last 3 days so far.

 

And its back. was gone, zero issues for multiple days. decided to play golf with some people and it decided that i can no longer do anything once again.

[Robchil]

9 hours ago, Axtasium said:

I turned off "turn on fast startup" and I haven't had an issue since. I'm not sure what exactly was going on, but since i turned that off and restarted my pc, there have been zero issues. seems like its some wack windows issue atm. Ill update this post if something does decide to flip out again but havnt had a single issue in the last 3 days so far.

 

And its back. was gone, zero issues for multiple days. decided to play golf with some people and it decided that i can no longer do anything once again.

what disk drives are you using. this can also be a failing drive. 

 

[Axtasium]

7 hours ago, Robchil said:

what disk drives are you using. this can also be a failing drive. 

 

That's a possibility i thought of. I'm using an external 5tb, an internal 2tb firecuda, and a 2tb sabrent rocket ssd. The only one that has any warnings on it is the firecuda one but its the same error that's been there for a long time now (before i started getting these errors). that number hasn't gone up at all in like a year and a half now. I Also went ahead and reseated my memory and gpu as I'm just going step by step so ill have to wait and see if that did anything. I will replace the hard drive at some point, probably with another ssd whether it be nvme or seta but its not something I want to do right now since the drive hasnt been having issues.

 

[Robchil]

2 minutes ago, Axtasium said:

That's a possibility i thought of. I'm using an external 5tb, an internal 2tb firecuda, and a 2tb sabrent rocket ssd. The only one that has any warnings on it is the firecuda one but its the same error that's been there for a long time now (before i started getting these errors). that number hasn't gone up at all in like a year and a half now. I Also went ahead and reseated my memory and gpu as I'm just going step by step so ill have to wait and see if that did anything. I will replace the hard drive at some point, probably with another ssd whether it be nvme or seta but its not something I want to do right now since the drive hasnt been having issues.

 

i havn't used harddrives for 10 years.. but had similar issues and swapped install drive.. from ssd to nvme..  that solved it... i will assume another ssd would do the trick too. as it's the system drive that failed. 

ssd's and nvme's fails differently from harddrives. as they "repair" themselves by replacing memory from a backup storage it has until it runs out.. then it just dies in the end. 

hence.. by replacing a bad sector you have a driver would create issues like these too. 

any dll that starts with nv are most likely nvidia driver. and d3d11.. is direct x 11.. 

 

[Axtasium]

1 hour ago, Robchil said:

i havn't used harddrives for 10 years.. but had similar issues and swapped install drive.. from ssd to nvme..  that solved it... i will assume another ssd would do the trick too. as it's the system drive that failed. 

ssd's and nvme's fails differently from harddrives. as they "repair" themselves by replacing memory from a backup storage it has until it runs out.. then it just dies in the end. 

hence.. by replacing a bad sector you have a driver would create issues like these too. 

any dll that starts with nv are most likely nvidia driver. and d3d11.. is direct x 11.. 

 

yeah, I'm aware its the GPU driver that its related to, issue is that apparently it could be literally anything causing the issue whether it be drivers, storage, settings, hardware, etc. I'm just trying to narrow the options of what it could be down. I'm at a loss at this point on what to do. Reseating my gpu and memory didn't fix the issue so i guess ill try reducing my memory from 3600 to 3200? i mean I've had this computer for like 3-4 years now and that was never an issue before, but maybe something is starting to act up? I never had this issue on my 2060 and then i swapped to a 3080. Had zero issues for awhile, then i got this new monitor which is a MSI MPG321QRF-QD, then a bit after i got it, the issue showed up. I don't know if maybe its something with the monitor maybe? or perhaps that it doesn't like being paired with the tv I'm using as a secondary or something. apparently its a known issue by nvidia. I saw them list it as a known issue on one of the recent driver pages, but its been an issue apparently for many people since about 3 drivers ago. thought maybe it was the drivers but the weird thing is that it doesn't seem to go away when i swap to older drivers. On one old driver windows decided to auto install, it was actually giving me errors every like 5 seconds so, seems like its not the driver specifically. I could also maybe try swapping to windows 11 maybe or its possible a windows update broke something as well. again, so many things it could be that its really hard to narrow things out right now. I just want to be able to run something without worrying about it randomly crashing on me.

[Robchil]

1 hour ago, Axtasium said:

yeah, I'm aware its the GPU driver that its related to, issue is that apparently it could be literally anything causing the issue whether it be drivers, storage, settings, hardware, etc. I'm just trying to narrow the options of what it could be down. I'm at a loss at this point on what to do. Reseating my gpu and memory didn't fix the issue so i guess ill try reducing my memory from 3600 to 3200? i mean I've had this computer for like 3-4 years now and that was never an issue before, but maybe something is starting to act up? I never had this issue on my 2060 and then i swapped to a 3080. Had zero issues for awhile, then i got this new monitor which is a MSI MPG321QRF-QD, then a bit after i got it, the issue showed up. I don't know if maybe its something with the monitor maybe? or perhaps that it doesn't like being paired with the tv I'm using as a secondary or something. apparently its a known issue by nvidia. I saw them list it as a known issue on one of the recent driver pages, but its been an issue apparently for many people since about 3 drivers ago. thought maybe it was the drivers but the weird thing is that it doesn't seem to go away when i swap to older drivers. On one old driver windows decided to auto install, it was actually giving me errors every like 5 seconds so, seems like its not the driver specifically. I could also maybe try swapping to windows 11 maybe or its possible a windows update broke something as well. again, so many things it could be that its really hard to narrow things out right now. I just want to be able to run something without worrying about it randomly crashing on me.

well.. a new drive isn't too expencive and something you will be able to enjoy anyway if that wasn't the problem.. altho.. you still wouldn't realy enjoy the pc. but when the problem gets found. And.. you have tried about everything else.. drives do fail over time.. ram does too but hardly as often as disks when it has ran stabile for a while. and it would be a waste to swap if it's the drive.. a drive you can use anyway. Not saying you need to buy one, ask if someone you know have one spare so you can test 

 

 

[Axtasium]

3 minutes ago, Robchil said:

well.. a new drive isn't too expencive and something you will be able to enjoy anyway if that wasn't the problem.. altho.. you still wouldn't realy enjoy the pc. but when the problem gets found. And.. you have tried about everything else.. drives do fail over time.. ram does too but hardly as often as disks when it has ran stabile for a while. and it would be a waste to swap if it's the drive.. a drive you can use anyway. 

 

 

yeah, I was gonna try and replace it soon so i guess ill look at that but I probably wont be able to grab one for a bit atm. I don't have a ton of money atm because of medical issues and lots of work that is currently being done on my kitchen. only thing i can think of is to just unplug it and see, but i obviously use the drive so its a bit of a pain.

[Robchil]

Just now, Axtasium said:

yeah, I was gonna try and replace it soon so i guess ill look at that but I probably wont be able to grab one for a bit atm. I don't have a ton of money atm because of medical issues and lots of work that is currently being done on my kitchen. only thing i can think of is to just unplug it and see, but i obviously use the drive so its a bit of a pain.

is your system installed on the SSD or the HDD? 

 

[Axtasium]

5 hours ago, Robchil said:

is your system installed on the SSD or the HDD? 

 

The SSD, i would never put it on an HDD. Just tried to turn the ram down to 3200, didn't work either. apex keeps crashing in the menu after like 2 minutes. Ive burned through every option I have atm. Ive tried everything on my list. only thing left is hardware, or maybe the windows update is broken. checked it for issues and there weren't any so maybe i should just update to windows 11 and see at this point.

[Axtasium]

Oh, and here's to give you an idea of how far back I've gone driver wise and its still an issue. This driver is from July 2022, almost an entire year ago, and the issue is still here. it started recently and its never been an issue before. The most recent things i did was get a new monitor, update my drivers, and update windows.

[Robchil]

3 hours ago, Axtasium said:

Oh, and here's to give you an idea of how far back I've gone driver wise and its still an issue. This driver is from July 2022, almost an entire year ago, and the issue is still here. it started recently and its never been an issue before. The most recent things i did was get a new monitor, update my drivers, and update windows.

if you think the windows installation is broken..  boot into troubleshooting, go into advanced.. and somewhere in there you will find uninstall last update. 

try to uninstall last feature update. 

 

[Axtasium]

16 hours ago, Robchil said:

if you think the windows installation is broken..  boot into troubleshooting, go into advanced.. and somewhere in there you will find uninstall last update. 

try to uninstall last feature update. 

 

yeah, issue is i know I've done a feature update or 2, but it was there before that i think because i remember hoping it would fix it. I'm not really sure until windows pushes another full update whether that's the issue or not. I directly contacted Nvidia support to see if they had any advice or maybe if its a know issue atm. I ran prime95 for almost 2 hours and did memtest86 and it passed both cycles i did so it doesn't seem to be a hardware issue atm unless the gpu itself is having issues, but I don't think that's the case either. right now I'm kind of stuck until maybe an update gets pushed or something. only thing i can think to try atm is updating to windows 11 but its a pain and is kind of a last resort.

[Robchil]

44 minutes ago, Axtasium said:

yeah, issue is i know I've done a feature update or 2, but it was there before that i think because i remember hoping it would fix it. I'm not really sure until windows pushes another full update whether that's the issue or not. I directly contacted Nvidia support to see if they had any advice or maybe if its a know issue atm. I ran prime95 for almost 2 hours and did memtest86 and it passed both cycles i did so it doesn't seem to be a hardware issue atm unless the gpu itself is having issues, but I don't think that's the case either. right now I'm kind of stuck until maybe an update gets pushed or something. only thing i can think to try atm is updating to windows 11 but its a pain and is kind of a last resort.

not sure if that will help any but your choice..  just remember to check that your boot disk need to be GPT .. not MBR.. if it's not already.. if it UEFI boots already that shouldn't be a problem. 

 

[Axtasium]

4 hours ago, Robchil said:

not sure if that will help any but your choice..  just remember to check that your boot disk need to be GPT .. not MBR.. if it's not already.. if it UEFI boots already that shouldn't be a problem. 

 

Yeah, I went ahead and pulled out my memory sticks and tested each one, which no issues at all while running either stick alone, then stuck them back in and the issue is gone again. Sadly, that doesn't mean it wont be back because it always seems to come back but at least my I can run stuff for a day or 2 for now. For right now im just gonna wait and see what Nvidia support says and wait and see if it decides to come back again.