Killing Floor 2 randomly gets blue screen since I installed RTX 4090

[Sam]

My 3080 was fine, but since I installed 4090, Killing Floor 2, specifically, keeps occasionally getting blue screens and crashing (some times blue screens, sometimes crashes where I submit the report every time).

And yes, nvidia driver is latest version.

I ran a stress test with FurMark on the 4090 card for a long time and it seems to be fine. Not a single issue happened. The temperature peaks at around 70C. It's not overclocked or anything. I'm running it on standard mode (in fact I'm too old for overclocking and its issues... I just want a stable gaming computer for when I get some time to play). Now this... Any idea what could be causing this?

[Leon]

you may need to DDU the gpu driver to remove any previous gpu driver files and check again.

[Aya]

>unreal engine 3 in 2022

[Sam]

Leon původně napsal:

you may need to DDU the gpu driver to remove any previous gpu driver files and check again.

I did a clean install when I first installed it.

[Aya]

See above answer

Pumpkin Spice Iced Capp původně napsal:

>unreal engine 3 in 2022

[Leon]

Sam původně napsal:

Leon původně napsal:

you may need to DDU the gpu driver to remove any previous gpu driver files and check again.

I did a clean install when I first installed it.

what is the blue screen error that you get if u can remember?

[ZOD]

https://www.resplendence.com/whocrashed

[Sam]

Leon původně napsal:

Sam původně napsal:

I did a clean install when I first installed it.

what is the blue screen error that you get if u can remember?

Some watchdog violation.

[Sam]

ZOD původně napsal:

https://www.resplendence.com/whocrashed

Thanks. I did the analysis there and this is what I got from the whocrashed program (it's repeated multiple times, all the same):

```
Bugcheck code: 0x133(0x1, 0x1E00, 0xFFFFF800054FB320, 0x0)
Bugcheck name: DPC_WATCHDOG_VIOLATION
Driver or module in which error occurred: (0x00007ffd`1281ab84)
Bug check description: The DPC watchdog detected a prolonged run time at an IRQL of DISPATCH_LEVEL or above. This could be caused by either a non-responding driver or non-responding hardware. This bug check can also occur because of overheated CPUs (thermal issue).
Analysis: This is likely caused by a hardware problem, but there is a possibility that this is caused by a misbehaving driver.
This bugcheck indicates that a timeout has occurred. This may be caused by a hardware failure such as a thermal issue or a bug in a driver for a hardware device.
Read this article on thermal issues
A full memory dump will likely provide more useful information on the cause of this particular bugcheck.
```

[Sam]

Here's more information from windbg:

```
DPC_WATCHDOG_VIOLATION (133)
The DPC watchdog detected a prolonged run time at an IRQL of DISPATCH_LEVEL
or above.
Arguments:
Arg1: 0000000000000001, The system cumulatively spent an extended period of time at
DISPATCH_LEVEL or above. The offending component can usually be
identified with a stack trace.
Arg2: 0000000000001e00, The watchdog period.
Arg3: fffff800054fb320, cast to nt!DPC_WATCHDOG_GLOBAL_TRIAGE_BLOCK, which contains
additional information regarding the cumulative timeout
Arg4: 0000000000000000

Debugging Details:
------------------

*** WARNING: Unable to verify timestamp for nvlddmkm.sys
...
```

It seems to be the nvidia driver...

To be clear, the blue screen happens after the game freezes for an extended amount of time (a few minutes). So this kind of makes sense (the timeout thing)... even though I have no idea what it means.

[NouH]

Hello Sam. Have you tried underclocking your GPU by 10% or more just to test? (Clocking speed and vram frequency) I've had driver timeouts before and this solved it for me. The problem itself wasn't on my GPU (or maybe even yours if this method works), If what I said previously works, then It's probably some compatibility problem between your current gig and your gpu. I tested my GPU on my cousin gig and it worked without any problem so it's probably some weird compatibility issue when used with the other components in your gig, which might also be a sign of defective hardware.

Give it a try at least, can't hurt.

[Sam]

NouH původně napsal:

Hello Sam. Have you tried underclocking your GPU by 10% or more just to test? (Clocking speed and vram frequency) I've had driver timeouts before and this solved it for me. The problem itself wasn't on my GPU (or maybe even yours if this method works), If what I said previously works, then It's probably some compatibility problem between your current gig and your gpu. I tested my GPU on my cousin gig and it worked without any problem so it's probably some weird compatibility issue when used with the other components in your gig, which might also be a sign of defective hardware.

Give it a try at least, can't hurt.

Thank you for the suggestion. I really appreciate the genuine advice and sentiment.

I'm gonna be completely honest with you. I could've bought the OC version of my 4090 and I didn't. I'm too old to mess with my graphics card... good old days are gone when I blew up a PSU with overclocking. I'm too old to overclock and underclock. I'm not super-encouraged to mess with my graphics card as I don't understand it and don't have the time to do it. I just work during the day and want to enjoy an hour or two of KF2 when I'm done.

Currently I'm trying DirectX 10 mode... maybe that helps... I don't know. Command line arguments are probably the furthest I'd mess with. Unfortunately, there's no bios update for my board to try. I'm trying whatever safe solutions there are... but would appreciate suggestions.

[Aya]

Nod your head and accept the consequences of a a 2016 game hitting a 2006 game engine's content ceiling and subsequently being bloated up with additional content for the next 6 years.

Whenever I say "unreal engine 3 in current year", I'm referring to how KF2 runs on an engine that first released when the Xbox 360 was considered "next gen", is no longer supported by Epic Games and NVidia, and the game has much more maps, weapons, and cosmetic items than what the engine is designed to handle, all while pushing the graphical limitations of UE3 to their fullest

Most you can do is hope that KF3 skips straight to Unreal Engine 5

[Sam]

Pumpkin Spice Iced Capp původně napsal:

Nod your head and accept the consequences of a a 2016 game hitting a 2006 game engine's content ceiling and subsequently being bloated up with additional content for the next 6 years.

Whenever I say "unreal engine 3 in current year", I'm referring to how KF2 runs on an engine that first released when the Xbox 360 was considered "next gen", is no longer supported by Epic Games and NVidia, and the game has much more maps, weapons, and cosmetic items than what the engine is designed to handle, all while pushing the graphical limitations of UE3 to their fullest.

Well... KF1 doesn't crash xD...

I'm gonna keep waiting for an answer. I love this game :-)

[Leon]

Sam původně napsal:

Here's more information from windbg:

```
DPC_WATCHDOG_VIOLATION (133)
The DPC watchdog detected a prolonged run time at an IRQL of DISPATCH_LEVEL
or above.
Arguments:
Arg1: 0000000000000001, The system cumulatively spent an extended period of time at
DISPATCH_LEVEL or above. The offending component can usually be
identified with a stack trace.
Arg2: 0000000000001e00, The watchdog period.
Arg3: fffff800054fb320, cast to nt!DPC_WATCHDOG_GLOBAL_TRIAGE_BLOCK, which contains
additional information regarding the cumulative timeout
Arg4: 0000000000000000

Debugging Details:
------------------

*** WARNING: Unable to verify timestamp for nvlddmkm.sys
...
```

It seems to be the nvidia driver...

To be clear, the blue screen happens after the game freezes for an extended amount of time (a few minutes). So this kind of makes sense (the timeout thing)... even though I have no idea what it means.

nvlddmkm.sys is surely the nvidia driver, in this case there isn't much that can be done, since there isn't any other version of gpu driver for the 4090, you will most likely need to contact support at nvidia, or send ticket to TWI.

This might be a long shot to try and see if anything can be narrowed down or not, but... try these steps.

- Create system restore point, this way if the following steps didn't change anything, you can just revert back to this point where nothing is changed.
- press win + R
- type msconfig, and hit enter
- follow the steps on this pic https://i.imgur.com/Q1v4WcQ.png , basically go to services, then hide all microsoft services, disable all the services, ok, apply.
- restart the pc

play the game and see if you still ge thte BSOD or not, if yes, all you need to do is just go to the restore point you created and just load it to get everything back on, but if you don't get the bsod, you will have to find out which service is causing issue to your nvidia driver.

[NouH]

Sam původně napsal:

NouH původně napsal:

Hello Sam. Have you tried underclocking your GPU by 10% or more just to test? (Clocking speed and vram frequency) I've had driver timeouts before and this solved it for me. The problem itself wasn't on my GPU (or maybe even yours if this method works), If what I said previously works, then It's probably some compatibility problem between your current gig and your gpu. I tested my GPU on my cousin gig and it worked without any problem so it's probably some weird compatibility issue when used with the other components in your gig, which might also be a sign of defective hardware.

Give it a try at least, can't hurt.

Thank you for the suggestion. I really appreciate the genuine advice and sentiment.

I'm gonna be completely honest with you. I could've bought the OC version of my 4090 and I didn't. I'm too old to mess with my graphics card... good old days are gone when I blew up a PSU with overclocking. I'm too old to overclock and underclock. I'm not super-encouraged to mess with my graphics card as I don't understand it and don't have the time to do it. I just work during the day and want to enjoy an hour or two of KF2 when I'm done.

Currently I'm trying DirectX 10 mode... maybe that helps... I don't know. Command line arguments are probably the furthest I'd mess with. Unfortunately, there's no bios update for my board to try. I'm trying whatever safe solutions there are... but would appreciate suggestions.

Underclocking your GPU is very safe, the silicon in your GPU can only be damaged by thermal stress, which is often caused by high temperatures and overclocking, which we aren't doing and underclocking puts less load on your gpu. It's very easy to do, you go to your GPU settings in the gpu app (That comes after installing the driver), then you go to adjustment tab (if it's called that for your gpu), and you'll get a warning message, that can be disregarded because it's a warning for people that plan to OVERCLOCK without knowledge. There's no danger if you're going to underclock. Then you adjust Clock speed and vram frequency to -10%.

Believe me when I say that I've tried every single thing in the book to solve Driver timeout issues and none worked. Except this, so it's up to you really. Underclocking is very safe and you can easily revert it back to normal since it's just some settings.