Next - it seems your Windows error reporting is stuffed full of Bitdefender errors. Can we temporarily uninstall Bitdefender. To do that you would want to use the tool provided by Bitdefender.
Windows Defender will take over the instant that is uninstalled. Confirm that it has. Should be an icon by your clock - looks like a shield.
Okay, it’s uninstalled successfully.
The only other thing I see is Windows updates are a bit behind. Windows doesn’t install some things unless you manually start the installation. Type Windows Update in the Search box. Select “check for updates.” On the screen that opens it will likely say “You’re up to date”. Ignore that and press the button below “Check for Updates”.
Be prepared for a bit of computer downtime while it installs.
After you get Windows up to date give things a test. If you still crash let’s reinstall the AMD using the DDU tool to uninstall and clean the system.
https://www.wagnardsoft.com/forums/viewtopic.php?f=5&t=1490
Basically you extract the uninstaller from the download. Then unplug the Ethernet or disable the Wifi to avoid Windows automatic driver installation.
I like to boot the computer into safe mode then run the DDU tool. (This isn’t essential and you will need to confirm you have the windows password before you do so. If you are using a PIN it will ask for the actual password after running safe mode.)
On the tool you select Graphics type then AMD. Run the recommended “Clean then restart”.
(If Nvidia graphics were installed at one time you would want to run the tool again and clean those remnants out)
To add one thing on to this, the behavior you’ve been describing sounds like a textbook CPU overheat. It’s odd that your temps seemed so low while playing. I am wondering how your PC performs under a high load, since we’re otherwise going to have problems finding the issue.
I recommend a stress test to try to duplicate it without the game running. We’re going to try checking for overheating on the CPU or GPU by running a stress test. For this, I like to run three programs - HWMonitor for hardware sensing, Heaven Benchmarking Tool to test the GPU, and Prime95 to stress test the CPU.
You’re going to need to download these three programs, and run all of them simultaneously. For Heaven, try running in windowed mode at the high graphics setting. For Prime95 you want to run the “Blend” test. All of these need to run for 2-3 hours, so I recommend doing it at a time when you can casually monitor your HWMonitor results and keep an eye out for extended periods of overheating - like during a day you’re cooking something or watching a movie or something.
If you find overheating, you can take the computer to a PC tech to have that resolved. If the computer crashes or restarts spontaneously before the overheating test is over, you will need to do the same because there’s probably a power problem or an even more dangerous overheating issue. If you can run the test for more than 2-3 hours without an overheating CPU/GPU, then get some screenshots of the test and send them in to us. To do this:
- Maximize the HWMonitor window and expand all the nodes on the left
- Scroll all the way up
- Take a screenshot with the Print Screen (prtscn) key
- Open up the program Paint and press ctrl+v to paste in the test
- Crop the screen if you want to only show the test, then save it as Test1.JPG
- Scroll all the way down, then repeat steps 3-5.
- Upload them somewhere like imgur and link us to the results. We’ll use those to look for more options.
Let us know how it looks!
Yeah it’s currently installing updates. Ugh, it’s taking forever.
I’ve already did a DDU but would you like me to do it again?
No if the install was with DDU that would suffice. That update could take awhile - I warned you.
One other thing comes to mind. Test with DirectX12, DirectX 11 and the new DirectX 11 Legacy settings. Menu > System > Advanced pane.
I’m heading off for a bit. I’ll check this later Galthok
1 Like
I’ll report back later, Tratt.
Thank you for your patience and guidance.
https://postimg.cc/gallery/2x23pdn2y/
Okay, here is a link to my screenshots of my test results. I let the stress test run for about 2 and a half hours. Unfortunately, the GPU test had stopped when I checked on it but you can still see how hot it got.
Something is not right. The sensor is actually reading 72C in the tests. The sensor and the package should normally only be 10 C different.
Can you download and install AMD overdrive. And use it to check the CPU temp. It gives a more accurate reading
https://www.amd.com/en/technologies/amd-overdrive
Yes that CPU temperature does look much too high now. (Iooks like I need to apologize to Jessail here)
Okay, I’ve downloaded AMD Overdrive. What am I looking for?
you will want to install it. There is a temperature reading for the CPU its been a while since I have used it. repeat the tests watching the temperatures this app reads.
Okay, I’m going to have to run them today after work. Is there anything fixable that would be causing this problem of there being this much a difference between the package temp and CPU?
My CPU is being cooled by a liquid radiator if that helps at all. Maybe it’s time for a cleaning? That’s what my IT buddy suggested.
After looking at the temps the AMD Overdrive provides, it only shows the thermal margins. That is, how much below or over the temperature is to the max.
All In one or a custom loop ? If the radiator is working properly it should never have gotten that hot.
It’s an all in one.
I have to admit, I have no idea when the last time I cleaned it was. He suggested a clean, flush, and refill.
EDIT: Just idling my package temps hover around 10C while my CPU hovers around 30C.
That using the over drive utility?
That’s using HWMonitor. Using AMD Overdrive all I can see is my CPU cores are running 60C below the max.
Wait, I think I’ve figured out where to find my CPU and my processor temps. Let me send a screen shot.
Its not really accurate unless its under load. 30C idle is OK. I would service the water cooler the temps during the test were scary. HW is doing something weird with your CPU and package should always be within 10 C if the cooling is working correctly.
https://postimg.cc/cvtFz6rG
I’m guessing the first 7 numbers on the right are my package temps while the one directly below that is my CPU.
Those temps are saying that it’s that far BELOW the max.
Should i run the GPU stress test with the CPU to get an accurate reading or is just running the CPU required?
I’m going to have to run it today after work.
I get off of work at 1PM Central so expect an update at around 3 or 3:30 central.