Unreal Engine is exiting due to D3D device being lost. (Error: 0x887A0006 - ‘HUNG’)
I can play Conan for hours upon hours with my old PC setup without any issues on Low settings
Radeon RX 570 Series
Intel-R i5-4570 3.20GHz
16 GB RAM
Windows 10
Desktop
While on my setup which I bought brand new a little over a year ago, the game crashes even on Low settings giving me the D3D error.
Nvidia RTX 3060 Ti
Intel-R i5-11400 2.40GHz
16 GB RAM
Windows 10
Desktop
Things I have tried:
Re-installing/Updating Graphics Drivers
Re-installing/Updating Microsoft Visual C++ Redistributables
Re-installing DirectX Drivers
Installing Windows Updates
Disabling/Enabling Graphics Card in Device Manager(My PC does not have Integrated Graphics)
Verifying Integrity of files on Steam
Disabling Antivirus Software
Overclocking,- Undervolting,- Running the game without AfterBurner
Running PC in Balanced and Performance Mode
Re-installing Windows
Updating BIOS
Installing Epic Games launcher along with the latest Unreal Engine which also crashes upon launching with a similar D3D error.
Running the game Fullscreen, Windowed Borderless, Windowed, Low, Medium, High, Ultra settings.
Running tests on CPU, RAM and GPU for faults (Everything seems to be working normally)
It seems that the error has gotten more frequent, a couple of months ago I was able to play for hours before crashing, now it has gotten so bad that I can’t play at all anymore.
Before the error stated in the title I was getting the following error while playing the game:
[File:C:\UE4\CSAND\Engine\Source\Runtime\RenderCore\Private\RenderingThread.cpp] [Line: 959]
GameThread timed out waiting for RenderThread after 60.00 secs
Max Power 500W
Airflow is definitely an issue with this HP Pavilion TG01.
I will have to check the BIOS…
RAM speed is 3200MHz.
I have always played at 1080p, 1920 x 1080 without any issues, I have over 1000 hours in Conan with this PC, had no issues at all before August of 2023.
okay I wasn’t sure there would be as that cpu doesn’t support overclocking. These d3d/renderthread errors are always a problem because its a base unreal error that occurs to protect hardware as something goes wrong. Obviously, it’s not an easy thing to change a PSU especially if you aren’t certain it will fix the problem (but it’s a strong suspect because a 3060 Ti recommends a 650W at minimum and I’m a fan of always getting larger so as not to tax the supply too much), but that and overheating would be the first things I would check. Have you tried simply taking off the sidewalls to your pc and seeing if just getting cooler air across everything makes a difference?
Yes I have removed the side panel for now for extra airflow, it does run slightly cooler but that doesn’t seem to be the culprit.
It’s very strange because if for example I join a new server and start off in the desert where there isn’t a lot to load in, the game runs really well. I recently just played on a server where I started from scratch, for weeks without any issues. Once I have built and gathered plenty of materials and built a decent size base the game begins to crash with the error ‘‘GameThread timed out waiting for RenderThread after 60.00 secs’’. Eventually when I have gotten that error plenty of times, it turns in to the D3D error…
have you gone into task manager and monitored your resources to see what runs up to max leading up to this? The origin of the error is that the gpu memory starts to fill and then it drops the device driver to save the hardware. I doubt tho that it’s actually filling the memory which leads me to suspect the PSU just can’t provide enough power. 500 W is pretty slim and beneath what’s recommended for that card and what you have tried so far pretty much leaves this as the natural next thing to address.
The fact that it starts the crash when you build is not really all that surprising as that’s when players take control of how much resources the game takes. If the whole map was layered with foundations and torches, it would probably crash even the best PC.
I have tried to monitor it through the Task Manager but when the crash happens, I can’t do anything on my desktop until the error actually appears, by that time the game has already shut down.
I was looking at this and wanted to ask, you upgraded to a 3060 from the stock 1650 correct? Did you try putting back the 1650 to see if the crash stops? This would also narrow it down to the PSU simply being too weak.
I have been considering buying a used GPU just to test it out to be honest but at the same time that feels like a waste of money. Better idea would be to buy a new case, motherboard, power supply, CPU and RAM. Something more compatible with the RTX 3060 Ti.
Right now it’s just a huge bottleneck and it’s not possible to upgrade much with the hardware it has currently.
It’s truly very odd and does not seem like hardware related, I was just able to play for at least two hours without any problem(my computer was chilling and not working hard at all with some being medium and some high settings) until I pointed my character at a specific location, the game freezes, shows the intro video image exactly where I skipped it from and back to desktop with D3D error.
Again, that’s not really surprising because when you changed the view, it needed to render things, some close, some far. In unreal engine with open world games, there is a process known as level streaming which loads levels into your client to view as you approach them/need to see them. The levels can be light aka not much to show or quite heavy if there is a ton of wildlife, foliage, structures, particle effects, etc. Where you are on the map will vary substantially in the amount of work your PC/Graphics card must do. Again, this is not a conan thing; this is a function of how Unreal Engine works. It makes perfect sense that if the problem is that your PSU simply cannot provide enough power to the GPU, this could happen and be variable depending on graphics load. It doesn’t mean necessarily that a new PSU will solve all your problems, but 500 W is beneath the recommended power per NVidia for that card, so having graphics issues is to be expected.
Still doesn’t explain how I managed to play over a 1000 hours without getting a single crash before. Everything worked fine until Age Of War. It does make sense that the game capacity is growing with each update and the hardware requirements become more…
by using the config option, but a stressed PSU will become damaged so, yes, it still holds it could be your PSU. It is simply not the recommended minimum power for your card. Regardless if the config option fixes the issue, you should upgrade your PSU.
Lots of people play the game with lesser systems without this error.
I tried using the -autoconfig option also before, same issue.
Like I mentioned in my initial post, I can play the game fine with my old PC which is running with:
Radeon RX 570 Series
Intel-R i5-4570 3.20GHz
16 GB RAM
I dismantled it today and found out that it has a 750W PSU.
Unfortunately the HP Pavilion TG01’s motherboard has a different connector and the PSU itself doesn’t fit in the case anyway. Even though my old PC’s CPU is seriously outdated and an even bigger bottleneck for the RTX 3060 Ti. I was thinking of swapping the Radeon RX 570 with the RTX 3060 Ti just to test how the game runs in my old machine.
I already did that, didn’t work. Also tried the variatons of installing only the Graphics Driver without GeForce Experience, with GeForce Experience, tried Game Ready Driver, Studio Driver, with Physics, without Physics etc…
We understand your frustration but we have to get the basics out of the way before starting an investigation. Now, could you send the log .txt file to us, please?
You can find it by following: Steam\steamapps\common\Conan Exiles\ConanSandbox\Saved\Logs