solved Asus RTX 3080 TUF crashes

[Aqua_HeaD]

Since I changed my GPU from my earlier RTX 2080Ti to my sons Ausus RTX 3080 TUF DCS crash every time randomly. It can happen within 5 minutes or after one hour. 

I have tried several different drivers from the old 472.12 up to the most recent 527.37 with the same result. I always use DDU in safe mode when I change drivers. My RTX 2080Ti has been working before I replaced it with the 3080 without any issues and also my old GTX 1080 work fine. 

I have tride to underclock/undervolt the 3080 but DCS still crash. 

My RAM memory has been checked without any issues and I have also checked my system SSD also without any problems.

I can run both Heaven Benchmark, Firestrike and Furmark without any crashes.

Any suggestions what can be wrong?

dcs.20221203-190329.log dxdiag.txt

Edited December 3, 2022 by Aqua_HeaD

[5ephir0th]

Have you try deleting FXO and Metashader folder from the saved games folder of DCS?

[Aqua_HeaD]

No I have not but I will do it immediately and see if it will be better.

 

[Flappie]

I've seen this crash before, here:

and here:

It looks like this ("DensityMapParams::CloudLayer" preceded by "memcpy"):

2022-12-03 19:03:28.916 WARNING DX11BACKEND (Main): Can't map structured buffer.
2022-12-03 19:03:28.916 ERROR   DX11BACKEND (Main): Failed assert `0` at Projects\render\dx11backend_win8\Source\DX11StructuredBuffer.cpp:100
2022-12-03 19:03:28.916 INFO    EDCORE (Main): try to write dump information
2022-12-03 19:03:28.920 INFO    EDCORE (Main): # -------------- 20221203-190329 --------------
2022-12-03 19:03:28.921 INFO    EDCORE (Main): DCS/2.8.0.33006 (x86_64; Windows NT 10.0.22000)
2022-12-03 19:03:28.922 INFO    EDCORE (Main): C:\WINDOWS\SYSTEM32\VCRUNTIME140.dll
2022-12-03 19:03:28.925 INFO    EDCORE (Main): # C0000005 ACCESS_VIOLATION at F841140A 00:00000000
2022-12-03 19:03:28.927 INFO    EDCORE (Main): SymInit: Symbol-SearchPath: '.;D:\Installerade spel\Eagle Dynamics\DCS World;D:\Installerade spel\Eagle Dynamics\DCS World\bin;C:\WINDOWS;C:\WINDOWS\system32;SRV*C:\websymbols*https://msdl.microsoft.com/download/symbols;', symOptions: 532, UserName: 'Jerker'
2022-12-03 19:03:28.929 INFO    EDCORE (Main): OS-Version: 10.0.22000 () 0x100-0x1
2022-12-03 19:03:28.948 INFO    EDCORE (Main): 0x000000000000140A (VCRUNTIME140): memcpy + 0x11A
2022-12-03 19:03:28.949 INFO    EDCORE (Main): 0x000000000002088E (enlight): enlight::CloudDetails::DensityMapParams::CloudLayer::isEnabled + 0x2DE
2022-12-03 19:03:28.949 INFO    EDCORE (Main): 0x00000000000F4832 (Visualizer): smSceneManager::regLua + 0x2222
2022-12-03 19:03:28.950 INFO    EDCORE (Main): 0x00000000000E6BFF (Visualizer): smSceneManager::DestroySceneManager + 0x435F
2022-12-03 19:03:28.951 INFO    EDCORE (Main): 0x000000000080E933 (DCS): SW + 0x3CEE43
2022-12-03 19:03:28.951 INFO    EDCORE (Main): 0x000000000082F2D9 (DCS): SW + 0x3EF7E9
2022-12-03 19:03:28.951 INFO    EDCORE (Main): 0x00000000007EE914 (DCS): SW + 0x3AEE24
2022-12-03 19:03:28.952 INFO    EDCORE (Main): 0x00000000007EF7E4 (DCS): SW + 0x3AFCF4
2022-12-03 19:03:28.952 INFO    EDCORE (Main): 0x0000000002289650 (DCS): AmdPowerXpressRequestHighPerformance + 0xF6064C
2022-12-03 19:03:28.953 INFO    EDCORE (Main): 0x0000000000BDCB52 (DCS): SW + 0x79D062
2022-12-03 19:03:28.955 INFO    EDCORE (Main): 0x0000000000015550 (KERNEL32): BaseThreadInitThunk + 0x10
2022-12-03 19:03:28.956 INFO    EDCORE (Main): 0x000000000000485B (ntdll): RtlUserThreadStart + 0x2B
2022-12-03 19:03:29.051 INFO    EDCORE (Main): Minidump created.

 

Can you please try a mission with no clouds at all (preset = "No Clouds") and see if you can play without crashing?

[Aqua_HeaD]

The game still crashed after about 25 minutes after I deleted the FXO and Metashader folder.

dxdiag.txt dcs.20221203-232809.log

[Aqua_HeaD]

Now it just crashed when I was about to start a mission. This is relly strange.

dcs.20221203-235253.log

[Special K]

vor 8 Stunden schrieb Aqua_HeaD:

Now it just crashed when I was about to start a mission. This is relly strange.

dcs.20221203-235253.log 512.98 kB · 5 Downloads

Your pagefile is way too low. That can lead to all kind of issues.

Decrease your preload radius to max 60000 and set your pagefile to at least 32 GB initial on your fastest SSD. Problems should vanish then.

[Hiob]

Have you monitored the temperatures of your system? Crashes after a certain amount of time passes could indicate a heat buildup issue.

[Aqua_HeaD]

1 hour ago, Special K said:

Your pagefile is way too low. That can lead to all kind of issues.

Decrease your preload radius to max 60000 and set your pagefile to at least 32 GB initial on your fastest SSD. Problems should vanish then.

Thanks. I will try that. 

1 hour ago, Hiob said:

Have you monitored the temperatures of your system? Crashes after a certain amount of time passes could indicate a heat buildup issue.

My GPU will stay around 70 degrees Celsius with max load. CPU temp will be way under that. If I run my chassis without the side glass the temp of the GPU will be ten degrees lower. 

[Hiob]

56 minutes ago, Aqua_HeaD said:

My GPU will stay around 70 degrees Celsius with max load. CPU temp will be way under that. If I run my chassis without the side glass the temp of the GPU will be ten degrees lower. 

[Aqua_HeaD]

The game will still crash with a larger pagefile and decreased preliad radius. I have opened up my chassi to get cool air inside and I have now tested all the RAM sticks alone in different slots and all sticks together but it still crashes. 

[Special K]

vor 2 Stunden schrieb Aqua_HeaD:

The game will still crash with a larger pagefile and decreased preliad radius. I have opened up my chassi to get cool air inside and I have now tested all the RAM sticks alone in different slots and all sticks together but it still crashes. 

Can you send another log please?

[Aqua_HeaD]

50 minutes ago, Special K said:

Can you send another log please?

I will do it maybe tomorrow. A little busy from now on.

[Aqua_HeaD]

7 hours ago, Special K said:

Can you send another log please?

210457: DCS Instant Mission Case 1 Recovery Landing with TeamSpeak on. 211850: Same mission but with TS off. First crasch after 2 minutes and the secons after 5 minutes.

dcs.log-20221204-210457.zip

dcs.log-20221204-211850.zip

Edited December 4, 2022 by Aqua_HeaD

[Flappie]

I think we need ED support team on this one. Please open a support ticket: log into DCS website, click "Support", then "Create a new ticket". Send the link to this thread and explain them that you're not the only one having this exact same crash.

In the meantime, can you retry using your former RTX 2080Ti without changing anything in your GPU driver. Does it keep crashing or not?

[Aqua_HeaD]

I will send a message to the support team. I have now changed the card back to my RTX 2080Ti and I can fly without any issues (At least for just over an hour before i quit)

I noticed somthing in the MSI Afterburner hardware monitior that I thought was a bit strange. The "Temp Limit" and "Power Limit" jumps up and down between 1 and 0 during the gaming. Maybe thats correct but dont remember that I saw that behaviour with the RTX 3080. Can anyone explain what that´s all about?  The first picture inside the cocpit is with my RTX 2080Ti and the second picture from MSI Afterburner  with the values from the RTX 3080 card at the same mission.
 

Edit: The support team answered that it probably had with the driver to do and recommended me to update to the most recent one. I have already the latest driver at the moment and had also tested the majority of the old ones without any progress. They also told me to reduce the clock speed and I have tried that as well but I will make another try. 

Edited December 5, 2022 by Aqua_HeaD

[Special K]

vor 4 Stunden schrieb Aqua_HeaD:

I will send a message to the support team. I have now changed the card back to my RTX 2080Ti and I can fly without any issues (At least for just over an hour before i quit)

I noticed somthing in the MSI Afterburner hardware monitior that I thought was a bit strange. The "Temp Limit" and "Power Limit" jumps up and down between 1 and 0 during the gaming. Maybe thats correct but dont remember that I saw that behaviour with the RTX 3080. Can anyone explain what that´s all about? 
 

Edit: The support team answered that it probably had with the driver to do and recommended me to update to the most recent one. I have already the latest driver at the moment and had also tested the majority of the old ones without any progress. They also told me to reduce the clock speed and I have tried that as well but I will make another try. 

 

Have you tried rolling back the driver yet? I have issues since anything newer than 522.

Remove it with DDU, to make sure, everything is clean an shiny and try 522 or an even older one.

[Aqua_HeaD]

34 minutes ago, Special K said:

Have you tried rolling back the driver yet? I have issues since anything newer than 522.

Remove it with DDU, to make sure, everything is clean an shiny and try 522 or an even older one.

I have tried almost every driver from the most recent to the old 472.12. I always use DDU in safe mode when changing driver. 

[freehand]

I have the Tuf 12gb 3080 version none Ti and had no problems on various drivers not sure if this is of any use to you.

Edited December 5, 2022 by freehand

[Aqua_HeaD]

48 minutes ago, freehand said:

I have the Tuf 12gb 3080 version none Ti and had no problems on various drivers not sure if this is of any use to you.

 

It’s so wierd and I have no more ideas what to test. The next thing are to buy new PSU, RAM and maybe a new GPU. I don’t want to just throw away money before I really know what causes those crashes since they only appear in DSC and not in Arma 3 or all those benchmarks and stress tests. 

[freehand]

19 minutes ago, Aqua_HeaD said:

It’s so wierd and I have no more ideas what to test. The next thing are to buy new PSU, RAM and maybe a new GPU. I don’t want to just throw away money before I really know what causes those crashes since they only appear in DSC and not in Arma 3 or all those benchmarks and stress tests. 

Friend of mine has the tuf 3080 10gb version and he as far as I know has had no problems he only runs stable version I run the beta will have a word with him tomorrow & see if he has had any problems but I sure he would of said.

[Flappie]

Have you tried resetting your Nvidia Control panel options yet? (Restore default)

And you can also enable Debug Mode to see if that helps (NVCP >About > Debug Mode).

[Aqua_HeaD]

12 minutes ago, Flappie said:

Have you tried resetting your Nvidia Control panel options yet? (Restore default)

And you can also enable Debug Mode to see if that helps (NVCP >About > Debug Mode).

I have not restored it but I have tried the debug mode. Still crash randomly.

[Special K]

It's strange that it works with the 2080 TI without issues. And with the same drivers settings and options. It's very unlikely the GPU, but that sounds more or less like it.. 

[Aqua_HeaD]

53 minutes ago, Special K said:

It's strange that it works with the 2080 TI without issues. And with the same drivers settings and options. It's very unlikely the GPU, but that sounds more or less like it.. 

I agree but I think it is still very strange that the GPU runs fine in other games and also overclocked in Heaven Benchmark, 3D Benchmark and Furmark for quit long sessions without any issues. That make´s me  belive it is software related. I will take my sons PSU (Corsair AX1200i) and his RAM and try them in my PC but I am almost sure it will have no effect since I have tried the 3080 for a short period in his PC with random crashes in some of his games. Valorant worked great in his PC but not Modern Warfare II for example.