[PC] DXGI_ERROR_DEVICE_HUNG

[Jettaz]

Ok so I've done a lot of googling to try and solve this error. It has been driving me insane and I'm at my wits end in trying to figure out a solution.

My PC specs:

AMD Ryzen 7 3700x
Gigabyte x570 Gaming x Motherboard
Gigabyte RTX 2070 Super WindForce OC
8GBx2 Corsair Vengeance 2133MHz RAM
Corsair RM750 - Gold Cert.
Note that I have recently replaced the PSU as I thought that was the suspect.

Things that I have tried:

1. Clean windows install. Totally fresh. Went all nuclear.
   
2. Updated BIOS to the latest. F30a. (Also have tried multiple different BIOS updates to get here. Stock bios was F11)
   
3. Tried installing latest drivers (452.06). No luck.
   
4. Tried installing older drivers. (446.14). No luck.
   
5. Installed the drivers using DDU and without Nvidia GeForce. No luck.
   
6. Turned on Nvidia debug mode to disable factory OC. No luck. Downclocked my GPU using MSI afterburner. No luck. I've tried various downlocked speeds and none work.
   
7. Lowered my graphics settings for the game + downclocked my GPU and I could play the game in windowed mode and managed to complete the first mission and got through the sewers with Kamala. Crashed here and it refuses to load again.
   
8. I've edited my registry with the TdrDelay set to 10 (decimal)
   
9. The biggest possible fix that doesn't work: I took the 2070 Super back to the store and got a brand new replacement. Still the same issues. It is definitely not the card at this point.
   
10. I've ran Memtest to check if my RAM was faulty and they are not the issue.
    
11. I've ran FurMark with everything maxed and it did not crash after 20 minutes.
    
12. I've ran Unigine Valley benchmark with no issues at all.
    
13. I've tried another PSU with no luck at all.
    
14. Also tried capping FPS at 30 with no luck.
    
15. Updated DirectX
    
16. Tried a different PCIe slot. No luck.
    
17. I'm not using a high res texture pack, its not even installed.
    
18. I'm using 452.06 DCH right now. I have tested older drivers as well but none are successful. 446.14 and 419.35.
    
19. I'm running Windows 10 Pro 2004. I wasn't even aware of the HAGS setting so all my testing were with it off. I have also just tested it with it on and no difference. Still the same error.
    
20. No temperature throttling at all when I was playing the game when it did decide to run for a bit. But now it refuses to even get into the campaign so they are all cold starts.
    
21. I have checked my power settings and I have AMD Ryzen High Performance selected with Link State Power Management set to off.
    
22. Tried using the standard versin of the nvidia driver 452.06 as i was using the DCH version before. No luck.
    
23. Tried editing my pagefile. Increased it drastically to 3 times the amount of my current RAM. No luck. https://imgur.com/a/Knapzo9

Some other things to note:
* In event viewer when I crash it states:

The description for Event ID 13 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.


At this point I think I can at least confidently eliminate the possibility that it is hardware related and its something to do with the drivers/windows update/the game. My old GPU 1080ti ran the beta fine and has never caused this much issue.

Any other ideas that I have yet to try that could possibly fix this issue?

[Jettaz]

22:14:24:415 (00005000) > [D3D] cdc::PCDX12GraphicsCommandListPool::GetCommandList: pDevice->CreateCommandList(pDM->GetNodeMask(m_node), m_CLType, *ppAllocator, nullptr, __uuidof(**(ppCL)), IID_PPV_ARGS_Helper(ppCL)) failed with HRESULT 0x887a0005
22:14:24:416 (00005000) > [Render] Node 0 memory stats
22:14:24:416 (00005000) > [Render] VMEM Budget: 7243 MB
22:14:24:416 (00005000) > [Render] VMEM Available: 6145 MB
22:14:24:416 (00005000) > [Render] SMEM Budget: 7500 MB
22:14:24:416 (00005000) > [Render] SMEM Available: 7178 MB
22:14:24:416 (00005000) > [D3D] ERROR! Device removed detected (0x887A0006: DXGI_ERROR_DEVICE_HUNG)
22:14:24:417 (00010652) > [DXGI] Present test failed with HRESULT 0x887a0005
22:14:27:258 (00005000) > [Crash] Memory statistics:
22:14:27:258 (00005000) > [Crash] Total RAM = 15.9 GB (16328 MB)
22:14:27:258 (00005000) > [Crash] Avail RAM = 10.1 GB (10334 MB)
22:14:27:258 (00005000) > [Crash] Total virtual memory = 131072.0 GB (134217727 MB)
22:14:27:258 (00005000) > [Crash] Avail virtual memory = 131056.6 GB (134201950 MB)
22:14:27:258 (00005000) > [Crash] Unhandled exception at address 0x000000004109233c (code: 0xc0000005): "Access violation"

[Hammerstein]

Same. Running an i7-9700K, a 1080Ti, 32GB Ram - everything with the latest drivers. I just set my TdrLevel to 0 and will give it another try.

13:53:40:813 (00018960) > [D3D] cdc::PCDX12GenericGPUBuffer::BeginReadBack: GetD3DResource(node)->Map(0, &range, &pData) failed with HRESULT 0x887a0005
13:53:40:813 (00018960) > [Render] Node 0 memory stats
13:53:40:813 (00018960) > [Render] VMEM Budget: 9457 MB
13:53:40:813 (00018960) > [Render] VMEM Available: 3661 MB
13:53:40:813 (00018960) > [Render] SMEM Budget: 14717 MB
13:53:40:813 (00018960) > [Render] SMEM Available: 14094 MB
13:53:40:813 (00018960) > [D3D] ERROR! Device removed detected (0x887A0006: DXGI_ERROR_DEVICE_HUNG)
13:53:40:817 (00000392) > [DXGI] Present test failed with HRESULT 0x887a0005
13:53:40:820 (00017624) > [D3D] cdc::PCDX12GenericGPUBuffer::AllocResource: m_pAllocation->GetResource(0)->Map(0, &range, &pData) failed with HRESULT 0x887a0005
13:53:40:820 (00013168) > [D3D] cdc::PCDX12DeviceManager::CreatePlacedResource: m_pD3DDevice->CreatePlacedResource(pHeap, HeapOffset, pDesc, InitialState, pOptimizedClearValue, riid, ppvResource) failed with HRESULT 0x887a0005
13:53:40:821 (00017624) > [Crash] Memory statistics:
13:53:40:821 (00017624) > [Crash] Total RAM = 31.9 GB (32704 MB)
13:53:40:821 (00017624) > [Crash] Avail RAM = 16.0 GB (16414 MB)
13:53:40:821 (00017624) > [Crash] Total virtual memory = 131072.0 GB (134217727 MB)
13:53:40:821 (00017624) > [Crash] Avail virtual memory = 130996.7 GB (134140649 MB)
13:53:40:821 (00017624) > [Crash] Unhandled exception at address 0x0000000041530bae (code: 0xc0000005): "Access violation"

Beta ran fine. All other hardware hungry games I play, like Division 2, also run fine. It's only in Avengers that I have this problem.

[Hammerstein]

Could play for 2+ hours with TdrLevel=0 then had a hard crash and my whole system locked up. Likely caused by the game crash and TdrLevel set to 0, meaning the display driver could not recover.

So that was not the solution. I removed the line from the Registry again.

[John [UK]]

Getting this too. Really annoying. I reset my graphics back to default and restarted my pc. I actually managed to finish a mission. we'll see how far I get before it starts up crashing again

[Blackwolfe]

Originally posted by Hammerstein:

Could play for 2+ hours with TdrLevel=0 then had a hard crash and my whole system locked up. Likely caused by the game crash and TdrLevel set to 0, meaning the display driver could not recover.

So that was not the solution. I removed the line from the Registry again.

I got hard crashes without tdrlevel set to 0 so it's probably not that.

2 hours, lucky you. I can't even play for 30 minutes without the game crashing.

[Jettaz]

Just tried running my GPU in the other PCIe slot. No luck. I could boot the game, get into the main menu to select campaign but the moment the campaign loads I crash.

No error displayed but looking at the logs its still the same error.

16:39:57:043 (00011712) > [DXGI] Present test failed with HRESULT 0x887a0005
16:39:57:043 (00011712) > [Render] Node 0 memory stats
16:39:57:043 (00011712) > [Render] VMEM Budget: 7243 MB
16:39:57:043 (00011712) > [Render] VMEM Available: 5542 MB
16:39:57:043 (00011712) > [Render] SMEM Budget: 7500 MB
16:39:57:043 (00011712) > [Render] SMEM Available: 7168 MB
16:39:57:043 (00011712) > [D3D] ERROR! Device removed detected (0x887A0005: DXGI_ERROR_DEVICE_REMOVED)
16:39:57:045 (00011800) > [D3D] cdc::PCDX12DeviceManager::CreatePlacedResource: m_pD3DDevice->CreatePlacedResource(pHeap, HeapOffset, pDesc, InitialState, pOptimizedClearValue, riid, ppvResource) failed with HRESULT 0x887a0005
16:39:57:045 (00011800) > [Crash] Memory statistics:
16:39:57:045 (00011800) > [Crash] Total RAM = 15.9 GB (16328 MB)
16:39:57:045 (00011800) > [Crash] Avail RAM = 8.6 GB (8806 MB)
16:39:57:045 (00011800) > [Crash] Total virtual memory = 131072.0 GB (134217727 MB)
16:39:57:045 (00011800) > [Crash] Avail virtual memory = 131054.7 GB (134200026 MB)
16:39:57:045 (00011800) > [Crash] Unhandled exception at address 0x00000000414981a0 (code: 0xc0000005): "Access violation"

[TotesAdorbs]

That was good thinking on the TDR change. That was going to be the first thing I suggested.

Are you running the game with the High-Res Texture pack?

Also, are you using the standard version the Nvidia drivers or the DCH version?

Last question, which version of Windows 10 are you running? If it is 2004, what is your HAGS setting?

Hope I can help!

*EDIT*:

Last, last question, though this is likely unrelated, when you run GPU-Z, what does your PCI-E Link state report? Is it PCI-E 3.0 x16 or is it something different such as x8 or 2.0? Have you checked your GPU and CPU temps while running the game?

[Jettaz]

Originally posted by TotesAdorbs:

That was good thinking on the TDR change. That was going to be the first thing I suggested.

Are you running the game with the High-Res Texture pack?

Also, are you using the standard version the Nvidia drivers or the DCH version?

Last question, which version of Windows 10 are you running? If it is 2004, what is your HAGS setting?

Hope I can help!

*EDIT*:

Last, last question, though this is likely unrelated, when you run GPU-Z, what does your PCI-E Link state report? Is it PCI-E 3.0 x16 or is it something different such as x8 or 2.0? Have you checked your GPU and CPU temps while running the game?

Thanks for taking the time to suggest some solutions!

• I'm not using a high res texture pack, its not even installed.
  
• I'm using 452.06 DCH right now. I have tested older drivers as well but none are successful. 446.14 and 419.35.
  
• I'm running Windows 10 Pro 2004. I wasn't even aware of the HAGS setting so all my testing were with it off. I have also just tested it with it on and no difference. Still the same error.
  
• I've got my card slotted into a PCIe x16 3.0 and it does run at that speed.
  
• No temperature throttling at all when I was playing the game when it did decide to run for a bit. But now it refuses to even get into the campaign so they are all cold starts.
  
  I also should mention I have checked my power settings and I have AMD Ryzen High Performance selected with Link State Power Management set to off.

[Pekmez Pita]

So much for it being a 10 series card issue only. Let me guess, Nixxes? Nvidia's still to blame?
But of course!

They have not addressed one issue from their "betas". Not even one.

[TotesAdorbs]

Hmmm...
This is an odd one for sure. The only possible thing I could think of would be using DDU to completely clean and remove the drivers and then installing the Standard drivers instead of the DCH, I've had issues with the DCH in the past.

Other than that, not sure what the issue could be. I can tell you that on our processor, the 3700X (I have the same), it has it's own in-built PCI-E controller in addition to the controller on the X570 mother board, so if you're overclocking it in anyway, or your RAM (because our processors are weird like that, lol) then that could for some weird reason be affecting the PCI-E interface. Then again, I would think that it would show up on a stress test, so I'm not really sure.

Lastly, if you haven't already, make a post on the Nvidia official forums in the Driver feedback thread at geforce.com, and also at the Nvidia subreddit. Include all the information you've included in these posts (not just a link to these posts) and watch for replies. Often times there's a few helpful people there, far more knowledgeable than me. Keep an eye out for one named "Sora" (I think, my memory could be mistaken) as they often respond with helpful advice and/or requests for info. Also, the Nvidia devs will sometimes request more information to help track down an issue as well, they might like to have some logs from your system.

Hope this helps! Good luck!

[Jettaz]

Originally posted by TotesAdorbs:

Hmmm...
This is an odd one for sure. The only possible thing I could think of would be using DDU to completely clean and remove the drivers and then installing the Standard drivers instead of the DCH, I've had issues with the DCH in the past.

Other than that, not sure what the issue could be. I can tell you that on our processor, the 3700X (I have the same), it has it's own in-built PCI-E controller in addition to the controller on the X570 mother board, so if you're overclocking it in anyway, or your RAM (because our processors are weird like that, lol) then that could for some weird reason be affecting the PCI-E interface. Then again, I would think that it would show up on a stress test, so I'm not really sure.

Lastly, if you haven't already, make a post on the Nvidia official forums in the Driver feedback thread at geforce.com, and also at the Nvidia subreddit. Include all the information you've included in these posts (not just a link to these posts) and watch for replies. Often times there's a few helpful people there, far more knowledgeable than me. Keep an eye out for one named "Sora" (I think, my memory could be mistaken) as they often respond with helpful advice and/or requests for info. Also, the Nvidia devs will sometimes request more information to help track down an issue as well, they might like to have some logs from your system.

Hope this helps! Good luck!

Thanks for the help. I tried a solution that someone else suggested and managed to play for a whole 2 minutes!

I turned on full screen, set resolution to 1920x1080 and could finally load into a campaign mission. Unfortunately crashed soon after.

I've posted in the driver feedback thread and I'll make a reddit post as well. Fingers crossed theres something I'm stupidly missing.

[GoblinCheeksClapper]

I do not believe this is a Hardware related issue and it's a code handling issue. I dont know what happens in this particular moment (maybe there's an invisible trigger or flag?) but whatever the player is activating something must be conflicting. Bummer. I'll play online instead!

[Darío]

I have 1080 ti and 32 gb ram and crash in the
sewer part.

[MachJava]

https://www.nvidia.com/en-us/geforce/forums/game-ready-drivers/13/396210/announcing-geforce-hotfix-driver-45222-released-94/

Anyone try the newest hotfix driver from NVIDIA? I have the same issue as the OP, but no device hung error, just CTD with no errors randomly, when performance is fine for the most part.

[Blackwolfe]

Originally posted by MachJava:

https://www.nvidia.com/en-us/geforce/forums/game-ready-drivers/13/396210/announcing-geforce-hotfix-driver-45222-released-94/

Anyone try the newest hotfix driver from NVIDIA? I have the same issue as the OP, but no device hung error, just CTD with no errors randomly, when performance is fine for the most part.

I'm curious too. For me 446.14 worked though. Not gonna try the hotfix drivers until I know they work.