Aurora R7, display driver crashing, strange stuttering issues

Question

I have a really obscure issue that's been plaguing my PC for well over a year now, and any attempts to troubleshoot it has gotten me nowhere. I've also searched the internet extensively and can never find my exact issue, so it's been hard to narrow down any remaining solutions. So currently whenever I encounter any form of video playback for the first time after booting up my PC, my display will flash to a black screen and then recover. This gets recorded in Event Viewer as Event 4101, and states 'Display driver nvlddmkm stopped responding and has successfully recovered'. I cannot reproduce this crash for the remainder of that session, but if I restart my PC, open Chrome and try to play any kind of video (this could be ads playing on a website, and I've even seen it when booting up Steam if the front page is animated or has some kind of VFX in any way), then it will happen again. After this happens, if I open NVIDIA Control Panel, Low Latency Mode will be set to 'Custom' and will only allow me to select 'Off' and 'Ultra', when I previously had this set to 'On' before the crash. I also get different settings to choose from under Power Management Mode that weren't there before the driver crashed, which is really odd. Strangely enough it also switches to another NVIDIA High Definition Audio device and the one I start with on boot of Windows gets disabled. It's being output to the same monitor, so I'm not sure what's going on here. A weird thing that I've observed is that if I start a game before encountering any form of video playback, this crash cannot be produced afterwards. This was a consistent experience until recently, where I noticed playing a match of League of Legends beforehand does not prevent the issue from happening. But things like Valorant, Sea of Thieves, Halo Infinite and others do. But these games have their own issues, because it seems like I get stutters every time I am rendering a new asset for the first time. Typically the larger the asset is, the greater the stutter. Initial loads seem to always take longer than subsequent loads as well, but that hasn't always been consistent. Once these stutters happen the first time, it doesn't happen when rendering that same object again. It can only be reproduced if I exit the game and boot it up again. I can give game specific examples if necessary, but it's important to note that the hard crash I mentioned earlier has never happened during a game, and only from trying to play a video for the first time. I'm not sure if this issue is related, but it behaves similarly because it will happen the first time, every time. What I've tried: In this past month, I've done 3 separate clean Windows installations using 3 separate methods. I first did this in 2020 to fix a separate issue, and then I can't recall when this one started happening. First from the Windows media creation tool and booting from a USB, which didn't go well the first time as it seemed to freeze up on a black screen before finishing the installation process so I booted from the USB a second time. The second method was a Windows recovery tool (can't locate it, I forgot where this was on Microsoft's site), which I assume built a new Windows install from existing files. The problem with these two is that it didn't seem to actually delete everything, as a 39 GB Riot Games folder remained during both of these clean installations, and I don't remember something like this happening when I did my first one in 2020 off of the media creation tool, so it could be worth trying again. The third method I just used 'Reset This PC' and chose to remove everything and download from the cloud. This took 6-7 hours but seemed to remove everything. Windows downloaded a driver for my GPU right away, and the issue could be reproduced right after the reset was finished.I have used Display Driver Uninstaller many times during this past year, and none of the various latest drivers used during those times (or the current driver 511.23) have given me any different results. If I remove the driver using this method and then try to play a video again, the issue can't be produced. I assume integrated graphics is being used here but I can double check this again. I've also disabled all of the extra NVIDIA High Definition Audio devices, but the crash would still occur and then it would just default to Realtek Digital Output, which isn't being used by anything at the moment. When my headset is plugged in and in use, the crash doesn't affect it. But if Discord is open when the crash happens, it will ask me if I want to switch to an NVIDIA High Definition Audio device, so it's treated as if a new audio device is being plugged in.  I've tried plugging my DisplayPort cable into all of the other ports on my GPU, but that didn't solve the issue. Tried setting performance plan to High Performance and selecting 'Prefer Maximum Performance' in NVCP, but that also didn't prevent it from happening. Set Shader Cache Size to 100 GB as well thinking that could maybe help those stutters I see in games, but that hasn't solved anything either. Temps are normal as well (in and outside of games), GPU is sitting around 36°C and CPU around 32°C during normal desktop use when the crash occurs. At this point I'm assuming it's likely a hardware issue, maybe either with my GPU or PSU, but I don't have another GPU I could try and not sure how I would single out the PSU. I've tried monitoring GPU/CPU usage & power using MSI Afterburner in the past but I haven't done that in quite some time, and I'm not sure if any of that data could help. I can't recall all of the things I've tried, but these are the major ones that pop into my head as I've been trying to figure out the source of this issue for over a year now. I can monitor other things if needed or provide more system info, but any help or suggestions at all would be greatly appreciated as this issue is very obscure and hard to troubleshoot.  Specs: Display - ASUS PG279Q i7 8700k GTX 1080 16 GB DDR4 2666MHz 2TB 7200 RPM SATA 6Gb/s Alienware 850 Watt PSU - Also as an extra thing, this might just be a strange Windows bug and not at all related to this, but on every startup I get notifications from the Windows store that says it's starting these 4 packages: Microsoft visual C++ 2012 UWP desktop runtime package Microsoft visual C++ 2013 UWP desktop runtime package Microsoft visual C++ 2015 UWP desktop runtime package DirectX Runtime If I open the Windows store immediately and check the downloads section, I can catch these 4 packages appearing there but they come and go very quickly. I doubt this is at all connected to the issue I mentioned above because this wasn't happening before I did my clean Windows installations, and that display driver issue was still present, but thought I'd also mention this just in case. Feel free to ignore this part.

Tesla1856 · Answer

@Link2448 wrote:

I've done 3 separate clean Windows installations using 3 separate methods.

1. First from the Windows media creation tool and booting from a USB,

2. which didn't go well the first time as it seemed to freeze up on a black screen before finishing the installation process so I booted from the USB a second time.

1. This is the correct way.

2. If you can not clean-install Windows-10 Pro 64bit, with a quality/working flash-drive created with Microsoft.com Media Creation Tool

... perfectly, all the way through, including the First Time Setup

... you likely have some kind of hardware problem.

Try re-seating the video-card, or try the other slot . Run ePSA from outside of Windows (it must pass 100%).

Be sure all the fans are working and it is not over-heating. You can carefully use an air-compressor (and small brush) to remove all dust from fans and inside power-supply and especially the video-card and liquid-cooler radiator.

Vanadiel · Answer

I would suggest running pre boot diagnostics and see if any error codes show up.   Pre boot diagnostics

JodyOgden2 · Answer

I just posted about this same issue Im having after doing a clean install on my R6. Im thinking I have it narrowed down to the Intel drivers for my onboard graphics and chipset drivers for my processor. Im also running an Nvidia 1080Ti. When I uninstall the Intel Graphics GPU and drivers the problem seems to go away. I know there's known issues for intel drivers causing all kinds of things related to this.

Tesla1856 · Answer

@JodyOgden2 wrote:

When I uninstall the Intel Graphics GPU and drivers the problem seems to go away. I know there's known issues for intel drivers causing all kinds of things related to this.

I'm glad I don't have this problem.

I would not want to be depending on Intel to fix their software.

They make some good CPUs, and even their recent video-cards are not too bad ... but apparently, software is very hard to get right. Obviously, different teams.

Link2448 · Answer

Thanks for the replies everyone, it's helped out a lot.

So I ran ePSA from BIOS and all tests passed successfully without finding any issues. It asked if I wanted to run the full memory test, and that took about 3 hours and also ran successfully without any errors.

Then cleaned my PC thoroughly and reseated the GPU. Booted up and ran diagnostics test again (not the full memory test this time) and it came up fine. Boot to Windows and the crash was still happening.

I was about to test the other PCIe slot until I read JodyOdgen2's comment. Uninstalled Intel's display driver and the Intel Display Audio device, reboot, and then tried to reproduce the issue and there was no crash this time and everything was fine, and my mind was just blown. Interesting thing is this time the audio device I had on startup was the same device my PC switches to when the driver would crash. Can't believe this has probably been the issue the whole time. Problem is the drivers popped right back up in device manager afterwards, so how do I prevent Windows from downloading these drivers? Would it be enough to just disable them or something?

Thanks again by the way. If this is actually the source of the issue then that's just insane, and I'd be so happy to finally get rid of this.

JodyOgden2 · Answer

I also am pretty perturbed that this issue is still out there. I am glad you tracked it down though hopefully after a year they will get this sorted out but in today's world I'm sure we will have to find a workaround ourselves or buy a new PC to get it fixed. I've got a couple more things im going to try and If I succeed I will definitely update your post.

Iain K Mackay · Answer

I loved my R7, once I got rid of the Dell OEM GPU card. It continues to run flawlessly but it is a dust magnet and anyone that bought a non water cooled version must have fried something by now.

JodyOgden2 · Answer

https://www.groovypost.com/howto/prevent-automatic-driver-installation-in-windows-10/

Link2448 · Answer

I'd definitely be super interested to hear anything else you figure out, so thanks for that.

https://www.groovypost.com/howto/prevent-automatic-driver-installation-in-windows-10/

Unfortunately Windows still updated these devices despite doing this, so that's odd. I wanted to test all of this a bit further because it feels to me as if these drivers are conflicting with each other in some way given the behaviour of these crashes.

I used DDU to uninstall NVIDIA's drivers and reset my PC. This put me on "Microsoft Basic Display Adapter", and I tried to produce the crash and couldn't. Installed NVIDIA driver, reset again, and the issue could be reproduced. Interesting thing is that I have to restart my PC in these situations to produce the crash. After this I did the same thing I did last night and uninstalled Intel's display driver, reboot, and the crash could be produced despite the driver not being downloaded yet, although it was already in progress at that time so maybe that's why (and this was with the setting set to "No" for allowing Windows to automatically download drivers for my devices).

So something about this makes me feel like these drivers are conflicting in some way because getting rid of one or the other seems to solve the crashing issue specifically, or maybe it's something to do with the OEM card for whatever reason. Anyway, seems like the fix is not entirely consistent and there's more to it, but this is much closer to solving the issue than I've ever been so far. Despite that, I still can't really find posts online that detail this issue. So yeah, that also makes me think we probably won't be seeing this get fixed unless we try new hardware or get a new PC altogether. Really unfortunate and confusing, I don't want to troubleshoot any stuttering issues without getting this one solved first so I know it's not related. Everything is playable at least though, just really annoying.

Link2448 · Answer

I have thought many times that replacing the OEM card could solve all of these issues, but at this point I've spent so much time between 2019-2022 troubleshooting various obscure problems that I'd rather just not upgrade anything because I'm not sure I want to keep this system for much longer. Performance is generally speaking what I expected and sometimes exceeds it when you isolate it from the weird issues/bugs/stuttering/freezes I've experienced, but yeah I am not at all surprised that replacing the GPU improved your experience. My OEM card seems to be an MSI GTX 1080 Aero but just stripped of all of its labels or something, I dunno lol.

JodyOgden2 · Answer

What Ive read on the issue is that its a memory leak thats causing the issue. The fix was to download the Intel Graphics BETA Windows DCH Drivers from the Intel Download Center to fix these issues however I have done that also and it did fix a couple of my issues but I still have issues with certain applications so It didnt fix them all for sure

Link2448 · Answer

Installed Intel's beta drivers and it seemed to work pretty well for a while. First and second boot, no crash on two separate days. Restarted PC on the same day as the second boot, and the crash happened afterward. Restarted again to see if it was consistent and it didn't crash that time. Checked Windows Update a bit after this and it had already downloaded 27.20.100.9664 on its own despite having the beta drivers installed, so I guess there's no avoiding that one.

Seemed like it had solved the crashes on video playback at first but I guess it's going to try and update the beta drivers every time. Those same stutters I described still remained in games though.

At least the display issue seems to have been narrowed down to a few things, just hard to find a solution.

Alienware Desktops

Was this post helpful?