Question Seemingly random crashes ?

[coreybgg]

Hi everyone!

I built a new PC about a month ago and have been experiencing some crashes while gaming with this error being thrown:

dxgi_error_device_hung

I understand that this is related to the GPU hanging and Windows issuing a reset of the GPU driver.

Specs:

MB: AsRock B650 PG Lightning
CPU: AMD Ryzen 5 7600X
RAM: G.SKILL Trident Z5 Black RGB 32GB(2x16GB) DDR5 PC5-48000 6000MHz CL36 F5-6000J3636F16GX2-TZ5RK
GPU: Gigabyte RTX 4090 Gaming OC (Using 3x8 pin with the adapter included with the GPU)
PSU: MSI MPG Gold A850GF
OS: Windows 10 Pro
Storage: 1x NVME M.2, 1x SATA SSD, 1x 7200RPM HDD.

Initially, I got those crashes while playing Jedi Survivor and only at specific sections and upon checking steam discussions and forums, it seems like a lot of people a having the same with RT enabled, so I assume it's an issue with the game itself.

However, just a couple of days ago, I started playing Deus Ex: Mankind Divided and after playing for some time, the game will crash with the same error, which leads me to believe that it may be something hardware/driver related, even though some forums also have the same error present for this game.

I should note, that both games are using DX12. I researched further regarding the error specifically, not in relation to any game and found that there could be various reasons for this, one of them being RAM. Now, the RAM that I have is listed as supported on the AsRock MB page:

https://pg.asrock.com/mb/AMD/B650%20PG%20Lightning/index.us.asp#Memory

Though I am only seeing an XMP profile for 6000MHz in the BIOS which is what I'm using. Should there be an EXPO option available or is that irrelevant? I'm honestly not all that familiar with RAM and XMP profiles so I'm not sure what should be the case here. I did run a memory test from Windows which did not come back with any errors.

Another note, I did play the System Shock remake that released recently without any issues. 25 hours of playtime and only a single crash, which did not log the same error in the Windows event viewer. This game is using DX11, however.

What I have done to attempt and remedy this:

Used DDU to remove drivers and unplugged ethernet cable to ensure that Windows does not automatically install any.
Updated BIOS to latest version.
Reinstalled chipset drivers.

Initially, Deus Ex was crashing during the benchmark and after doing this, I was able to play for a couple of hours before the crash. I was also able to play Jedi Survivor for about 30 minutes without it crashing, where it would previously crash a few minutes in.

Is there anything else that I could check? Any other information that I should provide to be more useful or any BIOS settings to adjust?

Thanks!

Edit: Forgot to add, using MSI Afterburner to monitor temps. CPU has not went above 80C and is mostly at 70C. GPU rarely going above 60C even under load.

 

[DelenaСhelsiecain]

Since you've already tried updating your drivers, BIOS, and reinstalling chipset drivers, let's explore a few more possibilities.

First, make sure your GPU is properly seated in the PCIe slot and all power connections are secure. Additionally, check if your GPU temperatures are within normal ranges during gameplay. Even though your temps seem fine, it's worth monitoring them closely.

Regarding your RAM, the XMP profile for 6000MHz should be the correct one to use. The EXPO option might not be necessary in this case. However, double-checking your RAM modules' compatibility with your motherboard is a good idea.

To further troubleshoot, you can try running stress tests like FurMark or MemTest86 to check for any potential hardware issues. Also, consider checking for any specific game patches or updates that may address the crashes you're experiencing.

 

[coreybgg]

Hello,

Thank you for the quick response!

I just double-checked the GPU and I can confirm that it is properly seated. I'm also using the support bracket that was included with it as it is quite heavy. Cable connection is also seated properly.

Regarding the RAM, what I was wondering is if I should have the EXPO profile available, since I currently have only XMP even though the AsRock support list says EXPO is available. There are also only 2 profiles present, one at 4800MHz, which I'm not sure what the purpose of is, since the RAM does run at 4800MHz as default, and the 6000MHz one.

The games that I'm having issues with are up to date with latest versions.

The thing is, I can't really reproduce the conditions under which the crash will occur. It does not seems to be related to GPU/CPU/Memory load or temperatures, since I always have the Afterburner OSD enabled and I'm monitoring those.

I will look into checking memory with MemTest even though I already used the one that comes with Windows 10.

I would also appreciate any information about the XMP profile potentially causing the issue. I can also post some photos of the timings and voltage a bit later.

Edit: Also, based on the provided specs, do you think the 850W PSU is sufficient? The GPU Max TDP is 450W and the CPU is 110W.

 

[DelenaСhelsiecain]

Regarding your RAM, the XMP profile should be sufficient, and the EXPO profile might not be necessary. If your games are up to date and your monitoring shows no unusual load or temperatures, running MemTest can help investigate memory issues.

Your 850W PSU should be enough for your setup, considering the GPU's TDP and CPU's power consumption.

Make sure your GPU is securely seated and try stress tests like FurMark.

 

[Ralston18]

Look in Reliability History/Monitor and Event Viewer.

Either or both tools may be capturing some error codes, warnings, or even informational events just before or at the time of the crashes.

Reliability History is easy to use and the time line format may show some pattern of events.

Event Viewer requires more time and effort to navigate and understand.

To help with Event Viewer:

How To - How to use Windows 10 Event Viewer | Tom's Hardware Forum (tomshardware.com)

Any entries can be clicked for more details (e.g., error codes) but the details may or may not be helpful.

Also try running the built in Windows troublshooters. The troubleshooters may find and fix something.

Also use "sfc /scannow" and "dism".

References:

https://www.lifewire.com/how-to-use-sfc-scannow-to-repair-windows-system-files-2626161

How to use DISM command tool to repair Windows 10 image | Windows Central

 

[coreybgg]

Thanks for all the information!

I ran MemTest, it took quite a while but it came back without any errors. Though I did not see any reference to the memory clock being 6000MHz during the test, the XMP was loaded and when checking RAM info, it also showed 6000MHz, so I assume that there's nothing to worry about there.

Ran "sfc /scannow" and it did report some corrupt files that it was unable to repair, but I was able to repair those via "dism". Now sfc is showing that everything is good.

I also checked Reliability History and Event Viewer, not seeing any patterns in RH, but I did see some warnings in EventViewer that were logged soon after I booted up my PC:

The description for Event ID 2 from source NVIDIA OpenGL Driver 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: DrvSetContext failed functionality indeterminant (pid=6596 cncmd.exe 64bit) The message resource is present but the message was not found in the message table

As I understand, "cncmd.exe" is referring to Radeon Software, as the CPU has an integrated GPU. In fact, I did not use this at all before the DDU, so I guess this is why I'm seeing this for the first time.

Other than this, I'm seeing a lot of these soon after boot:

Reset to device, \Device\RaidPort0, was issued.

And this is what I got as the game crashed:

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 The message resource is present but the message was not found in the message table

And after the crash:

Display driver nvlddmkm stopped responding and has successfully recovered.

Also, I'm attaching images of the RAM config as mentioned earlier.

If someone more knowledgeable than me could let me know if they're seeing any issues. I'm using the XMP 1 profile and I'm wondering why there are EXPO 0 and EXPO 1 profiles that are staying empty. Though considering that MemTest did not throw any errors, I assume that RAM is not the issue.

As for changes, what I've done since the last crash is rolled back to oldest drivers available on Nvidia's website and repaired the packages using dism and sfc as suggested. These changes are yet to be tested, so I'll make sure to update once I've done so.

Cheers!

 

[coreybgg]

Well, I did manage to play Deus Ex for maybe 3 hours before it crashed with the same error. Still though, I'm not certain if this is related to the game itself, or something with the hardware. Any tips on what else I could check would be greatly appreciated!

 

[coreybgg]

Another update. After further researching I came across some information that having the PCIe Gen set on Auto in the BIOS could cause this issue. So I changed the PCIe Gen from Auto to Gen 4 in the BIOS and was able to play about 1 hour of Jedi Survivor without any crashes or driver errors. Had to stop (not due to crash) but I will test more again tomorrow and update the thread to confirm if this resolved the issue.

 

[coreybgg]

Well, after playing for a few days and hours on end without any crashes, last night, Deus Ex crashed twice within 2 hours with the exact same driver error.

Again, this doesn't seem to be related to the current load and temperatures and I'm not able to recreate the issue. One thin to note is that both crashes were when alt-tabbing during a dialogue, though I'm still unable to replicate.

Is there anything else I could try? Would it even make sense to try and change PCIe x16 to Gen 3 with a 4090? Will it have any significant performance impact?

 

[DelenaСhelsiecain]

Given everything you've tried, you might consider testing your system with a different GPU if possible. This could help identify if the issue is with the RTX 4090 itself.

Switching PCIe x16 to Gen 3 shouldn't have a significant performance impact, but it might be worth trying to see if it resolves the issue.

 

[coreybgg]

Thanks, I think I'll try switching to PCIe Gen 3 and see how it goes. Unfortunately, I don't have another GPU available to test if the issue will be present.

One thing that came to mind was modifying the TdrDelay in the Windows registry, but I'm not completely certain if doing so is safe. Can any GPU damage occur if Windows does not reset the driver in the default time when a hang is detected?

 

[DelenaСhelsiecain]

Modifying the TdrDelay in the Windows registry is generally considered safe in the sense that it's unlikely to cause physical damage to your GPU. What it does is it changes how long the system waits before it resets the graphics card driver when it's not responding.

However, please note that if you set this delay too long, it might make your system unresponsive for a longer period of time if the graphics card driver does hang. So, if you decide to go this route, you should proceed with caution and make sure you're comfortable with potentially needing to manually restart your system.

 

[MercadorHG]

Hello, I got the same error on Kingdom Come Delivrance, sometimes after 1-2 hour, sometimes after only 5 minutes. I got a similar build as yours :
PCPartPicker Part List

CPU: AMD Ryzen 5 7600 3.8 GHz 6-Core Processor ($309.00 @ Canada Computers)
CPU Cooler: Thermalright Peerless Assassin 120 SE ARGB 66.17 CFM CPU Cooler ($40.00)
Motherboard: ASRock B650M PG RIPTIDE WIFI Micro ATX AM5 Motherboard ($189.99 @ Newegg)
Memory: G.Skill Ripjaws S5 32 GB (2 x 16 GB) DDR5-6000 CL32 Memory ($104.99 @ Newegg)
Storage: Crucial P5 Plus 2 TB M.2-2280 PCIe 4.0 X4 NVME Solid State Drive ($192.99 @ PC-Canada)
Video Card: Asus DUAL GeForce RTX 4070 12 GB Video Card ($809.00 @ Canada Computers)
Case: Deepcool CG560 ATX Mid Tower Case ($89.99 @ Canada Computers)
Power Supply: Thermaltake Toughpower GF3 TT Premium 750 W 80+ Gold Certified Fully Modular ATX Power Supply ($149.99 @ Canada Computers)
Total: $294.98
Prices include shipping, taxes, and discounts when available
Generated by PCPartPicker 2023-07-10 16:44 EDT-0400

Have you found a way to fix this? I wonder if it's the DDR5 as I get some random errors on memtest, (like 1 to 4, never at the same place, only once per 4 runs) or the motherboard (currently running 1.24.AS02 as bios). Which bios version do you run? Thanks a lot!

 

[amcantech]

can you go through your windows event log and find any Errors that correspond roughly to the time of crash?

 

[MercadorHG]

can you go through your windows event log and find any Errors that correspond roughly to the time of crash?

It was the DDR timing that was causing the issues.

 

[Diegostarmo]

Hello, I got exactly the same type of crashes. Applications just suddently close without any crash message. This is my build:
CPU: AMD Ryzen 5 7600X
CPU Cooler: Cooler Master MasterLiquid ML240L V2
Motherboard: ASUS TUF Gaming B650-PLUS WiFi
Memory: Corsair Vengeance DDR5 32GB (2x16GB) 6000MHz C36
Storage: Kioxia Exceria Pro Unidad SSD 1TB NVMe M.2 2280 PCIe Gen4 x4
Video Card: Gigabyte GeForce RTX 4060 Ti AERO OC 8GB GDDR6 DLSS3
Power Supply: Corsair RMe Series RM750e 750W 80 Plus Gold Modular

I've seen that we both have the AMD Ryzen 5 7600x CPU, but it might just be a coincidence. If anyone has any update on this issue please let me know, thanks.

 

[MercadorHG]

Hello, I got exactly the same type of crashes. Applications just suddently close without any crash message. This is my build:
CPU: AMD Ryzen 5 7600X
CPU Cooler: Cooler Master MasterLiquid ML240L V2
Motherboard: ASUS TUF Gaming B650-PLUS WiFi
Memory: Corsair Vengeance DDR5 32GB (2x16GB) 6000MHz C36
Storage: Kioxia Exceria Pro Unidad SSD 1TB NVMe M.2 2280 PCIe Gen4 x4
Video Card: Gigabyte GeForce RTX 4060 Ti AERO OC 8GB GDDR6 DLSS3
Power Supply: Corsair RMe Series RM750e 750W 80 Plus Gold Modular

I've seen that we both have the AMD Ryzen 5 7600x CPU, but it might just be a coincidence. If anyone has any update on this issue please let me know, thanks.

Run memtest, my RAM at 6000mhz was causing the crashes. At base speed, no more crashs.