Rtx 4090 - games crashed upon first launch after download nvlddmkm

xxx123 · Dec 16, 2022

Hi. I have 12900K stock

32 GB 2x16 GB DDR4 3600mhz

Palit Gamerock Pro OC rtx 4090

1300W Seasonic Platinum Prime

Aorus Elite DDR4 Z690

Windows 11 2H22. Updated new nvidia drivers. Tried 4 versions of drivers the same issue.

I downloaded 3 games: Metro Exodus Enhanced,Watch Dogs Legion,Portal Rtx.
And on all of these 3 games i had crash on loading screen in my first launch ( after download ). Not happened again since. Now i can launching game forever and it will never crash. So why it crashed on first time? Thx

error log:

1.display driver nvlddmkm stopped responding and has successfully recovered

2.The description for Event ID 0 from source nvlddmkm cannot be found. Either the component that raises this event is not installed on your local computer or the installation is corrupted. You can install or repair the component on the local computer.

If the event originated on another computer, the display information had to be saved with the event.

The following information was included with the event:

\Device\Video3
Error occurred on GPUID: 100

Is my card fine?

Cromax · Dec 16, 2022

check power cables on card so that they sit right .. it seems it crashes on initial power draw during game start

xxx123 · Dec 16, 2022

Cromax said:
check power cables on card so that they sit right ..

they ok

Cromax · Dec 16, 2022

Did you use DDU uninstaller before you installed drivers?

Cromax · Dec 16, 2022

if you didnt go in safe mode clean gpu drivers with ddu, restart, boot in normal mode, install drivers and restart again

xxx123 · Dec 16, 2022

no i dont use ddu

Cromax · Dec 16, 2022

you should try it

xxx123 · Dec 16, 2022

I have an question.I have bios F7 motherboard. The newest is F21. ( 2 newer than me ). Is this worth to update or that doesnt help in future?

Improve the linkage between Resizable and 4G above
Add "Instant 6GHz" profile in CPU upgrade option (supports i9-13900K/KF and i7-13700K/KF)
Improve CPU OC stability, DDR compatibility & PCIe Gen5 stability

DigitalGoat · Dec 16, 2022

xxx123 said:
I have an question.I have bios F7 motherboard. The newest is F21. ( 2 newer than me ). Is this worth to update or that doesnt help in future?

Improve the linkage between Resizable and 4G above

Add "Instant 6GHz" profile in CPU upgrade option (supports i9-13900K/KF and i7-13700K/KF)

Improve CPU OC stability, DDR compatibility & PCIe Gen5 stability

The crashes on initial game launch could well be due to the shader compilation that happens during game launch, a newer driver might cure this but if the games now work fine it is something you can wait for Nvidia to get around to fixing, if indeed that is the issue.
A newer BIOS that targets stability issues and especially OC & graphics support in your case would be a sensible move, as long as it is not a beta version.
I agree with Cromax that using DDU in safe mode without internet is a good choice, just make sure you have downloaded the graphics package first, also while in safe mode look for DirectX shader cache folders (a list of locations should be found on Guru3d site forums) and delete the contents as DDU does not always do it.

xxx123 · Dec 16, 2022

DigitalGoat said:
The crashes on initial game launch could well be due to the shader compilation that happens during game launch, a newer driver might cure this but if the games now work fine it is something you can wait for Nvidia to get around to fixing, if indeed that is the issue.
A newer BIOS that targets stability issues and especially OC & graphics support in your case would be a sensible move, as long as it is not a beta version.
I agree with Cromax that using DDU in safe mode without internet is a good choice, just make sure you have downloaded the graphics package first, also while in safe mode look for DirectX shader cache folders (a list of locations should be found on Guru3d site forums) and delete the contents as DDU does not always do it.

ok thx so graphic card is propably fine?

clam1952 · Dec 16, 2022

xxx123 said:
ok thx so graphic card is propably fine?

Might want to check the power plugs are not melting which seems to be an issue with 4080 and 4090 cards.

DigitalGoat · Dec 16, 2022

If the plugs were melting then I doubt the card would crash only on first launch of each game then work fine.
The melting issue was blown out of proportion by sensationalist and click baity reporting with little to no research and testing.
Gamers Nexus has a video with a full technical breakdown and forensic laboratory results, backed up by info from Nvidia's own testing, and the results were that the plugs were not being seated properly by users, causing uneven heat build up.
The recommendation is to push the plugs fully in and not to bend the cable too severely especially near the plug.
Also the issue was being caused with the 3 & 4 PCI to 12 pin adapters supplied with the cards, 3rd party cards can be found with 8 pin PCI sockets or 3rd party 12 pin cables, but the same advice about socketing and bending applies to nearly all power cables.

For the OP, test the card using a burn in power virus like Furmark for temps & voltages, use Unigene Superposition to test VRAM and GPU. You can use MSI Afterburner and/ or HWInfo64 to monitor and log temps, voltages, core & memory clocks and a wealth of other info.

xxx123 · Dec 16, 2022

DigitalGoat said:
If the plugs were melting then I doubt the card would crash only on first launch of each game then work fine.
The melting issue was blown out of proportion by sensationalist and click baity reporting with little to no research and testing.
Gamers Nexus has a video with a full technical breakdown and forensic laboratory results, backed up by info from Nvidia's own testing, and the results were that the plugs were not being seated properly by users, causing uneven heat build up.
The recommendation is to push the plugs fully in and not to bend the cable too severely especially near the plug.
Also the issue was being caused with the 3 & 4 PCI to 12 pin adapters supplied with the cards, 3rd party cards can be found with 8 pin PCI sockets or 3rd party 12 pin cables, but the same advice about socketing and bending applies to nearly all power cables.

For the OP, test the card using a burn in power virus like Furmark for temps & voltages, use Unigene Superposition to test VRAM and GPU. You can use MSI Afterburner and/ or HWInfo64 to monitor and log temps, voltages, core & memory clocks and a wealth of other info.

So is this worth to clear nv shader cache? Can it help it?

That folder:"%username%\AppData\Local\Temp\NVIDIA Corporation\NV_Cache" or c:\program data\nvidia corporation\NV_Cache" or
"%username%\AppData\Local\Temp\NVIDIA Corporation\NV_Cache"

%PROGRAMDATA%\NVIDIA Corporation\NV_Cache\
%TEMP%\NVIDIA Corporation\NV_Cache\
%LOCALAPPDATA%\NVIDIA\GLCache\
%LOCALAPPDATA%\NVIDIA\DXCache\
%LOCALAPPDATA%\D3DSCache\

there's also COMPUTE cache:

%APPDATA%\NVIDIA\ComputeCache\

PS:
I tested Unigine Superposition and noo issues.

clam1952 · Dec 16, 2022

DigitalGoat said:
If the plugs were melting then I doubt the card would crash only on first launch of each game then work fine.
The melting issue was blown out of proportion by sensationalist and click baity reporting with little to no research and testing.
Gamers Nexus has a video with a full technical breakdown and forensic laboratory results, backed up by info from Nvidia's own testing, and the results were that the plugs were not being seated properly by users, causing uneven heat build up.
The recommendation is to push the plugs fully in and not to bend the cable too severely especially near the plug.
Also the issue was being caused with the 3 & 4 PCI to 12 pin adapters supplied with the cards, 3rd party cards can be found with 8 pin PCI sockets or 3rd party 12 pin cables, but the same advice about socketing and bending applies to nearly all power cables.

For the OP, test the card using a burn in power virus like Furmark for temps & voltages, use Unigene Superposition to test VRAM and GPU. You can use MSI Afterburner and/ or HWInfo64 to monitor and log temps, voltages, core & memory clocks and a wealth of other info.

Was just making the OP aware of a possible problem he may not have been aware of just in case.
Gamers Nexus was the first place I went to see what was actually happening and have followed all his subsequent videos on the subject so well aware of the exaggeration element.

xxx123 · Dec 16, 2022

Also have an question. The fact that the on RTX 3090 Ti i dont had this issue. So is my 4090 faulty or not? thx should i worry?

jds63 · Dec 16, 2022

xxx123 said:
Hi. I have 12900K stock

32 GB 2x16 GB DDR4 3600mhz

Palit Gamerock Pro OC rtx 4090

1300W Seasonic Platinum Prime

Aorus Elite DDR4 Z690

Windows 11 2H22. Updated new nvidia drivers. Tried 4 versions of drivers the same issue.

I downloaded 3 games: Metro Exodus Enhanced,Watch Dogs Legion,Portal Rtx.
And on all of these 3 games i had crash on loading screen in my first launch ( after download ). Not happened again since. Now i can launching game forever and it will never crash. So why it crashed on first time? Thx

error log:

1.display driver nvlddmkm stopped responding and has successfully recovered

2.The description for Event ID 0 from source nvlddmkm cannot be found. Either the component that raises this event is not installed on your local computer or the installation is corrupted. You can install or repair the component on the local computer.

If the event originated on another computer, the display information had to be saved with the event.

The following information was included with the event:

\Device\Video3
Error occurred on GPUID: 100

Is my card fine?

Had an issue very similar to yours as far as error codes in Event Viewer.
Freezes then restarts nvlddmkm.sys
Confused on how your exactly explaining your issue here as to all of sudden no issue with crashing anymore, when and what was done just before no crashing ? Was new driver installed for NVIDIA ? That as said could of cured it. Any BSOD or dump file ?

What does this mean ??? You download this file nvlddmkm.sys somewhere ?

“after download nvlddmkm”

Mine never crashed playing games i even stressed tested my card with Furmark no crash mine would freeze and get these errors when just clicking screen doing things on PC happen all of a sudden. So far mine has been good for moth, not sure my cure.

Solarstarshines · Dec 26, 2022

It's Nvidia's new drivers i am having the same issues with a 13900k went back to a older driver games worked but very few.
I feel like it's a platform thing i even went from a Z690 to a Z790 and still suffer for now i am just going to use my other rig for gaming till this problem is fixed it's pure BS but for now we are at the mercy of this driver :ffs:

DigitalGoat · Dec 26, 2022

Just to point out that Furmark is a torture test for GPUs and doesn't stress or test all the VRAM on a card, unless newer versions have changed.
The posts on this page may well be a result of buggy drivers (the last few iterations have certainly had thier fair share) but if there is an issue that only shows when the card is gaming then it is possible it could be caused by a faulty VRAM address.
Games may allocate all the available VRAM on a card but not use it all until specific sections of the game.
A benchmark like Unigene's Superposition (free) can be used to stress test all the cards VRAM using large texture sets (an error could lie in the upper 8th GB range of an 8GB card, for example) upto it's max.
The number of different hardware and software combinations make such troubleshooting extremely difficult of course, and the 40 series cards are not fully matured or publicly tested yet as far as drivers are concerned, making the issue harder to diagnose.

Solarstarshines · Dec 26, 2022

DigitalGoat said:
Just to point out that Furmark is a torture test for GPUs and doesn't stress or test all the VRAM on a card, unless newer versions have changed.
The posts on this page may well be a result of buggy drivers (the last few iterations have certainly had thier fair share) but if there is an issue that only shows when the card is gaming then it is possible it could be caused by a faulty VRAM address.
Games may allocate all the available VRAM on a card but not use it all until specific sections of the game.
A benchmark like Unigene's Superposition (free) can be used to stress test all the cards VRAM using large texture sets (an error could lie in the upper 8th GB range of an 8GB card, for example) upto it's max.
The number of different hardware and software combinations make such troubleshooting extremely difficult of course, and the 40 series cards are not fully matured or publicly tested yet as far as drivers are concerned, making the issue harder to diagnose.

Just to be clear my issue is with a intel Z790 Mb and a EVGA 3090 I swapped the Z690 out for the same problem
I also use 64GB of TridentZ Royal DDR4 3600mhz nothing is overclocked yet my Ryzen 5950x and 3080 10GB has zero issues with any of the games i tried to play but the 13900k is just making me feel like it's garbage and isn't worth the headache.
Crashes make zero sense i even took out a samsung 980pro because it was crashing and dying from the ridiculous BSOD's

clam1952 · Dec 26, 2022

Solarstarshines said:
Just to be clear my issue is with a intel Z790 Mb and a EVGA 3090 I swapped the Z690 out for the same problem
I also use 64GB of TridentZ Royal DDR4 3600mhz nothing is overclocked yet my Ryzen 5950x and 3080 10GB has zero issues with any of the games i tried to play but the 13900k is just making me feel like it's garbage and isn't worth the headache.
Crashes make zero sense i even took out a samsung 980pro because it was crashing and dying from the ridiculous BSOD's

Have you tried the 3090 in the Ryzen rig or the 3080 in the Intel? which is probably what I would try, to verify if it's the GPU / CPU combination that's got a problem, I must admit being sceptical about the Intel introducing effectively a two tier system of cores.

Anything on the Nvidia forums about similar problems?