Question AMD Driver Timeout is driving me crazy. Please help.

Jun 22, 2022
13
0
10
0
For a few months I've been experiencing a lot of driver timeouts, no matter the game with my RX 550.

It's not temperature because when it happens GPU temperature is not even hot and GPU use is low, Windows is up to date as well.

I get error Display 4101 on Event log with the description [Event ID: 4101-Display driver amdwddmg stopped responding and successfully recovered]

Things I've tried:
1. Using DDU to uninstall and install different drivers, I tried the oldest driver for my current OS, a random driver and the newest driver. None worked. I've also selected for windows to not install it's default driver (Thinking that could have been messing with it somehow, it didn't work either).

2. Different Harddisks/SSD with different OS installations. Didn't work.

3. Flashing my BIOS

4. Setting Virtual Memory to a manual value instead of automatic.

5. Clearing GPU shader cache.

I've tried quite a few other things as well which I can't remember right now, nothing seems to work. If anyone could help me at least identify the cause of this problem I would be grateful. Even if I can't solve it, just knowing what's causing it would help me a lot.
 
Tom's doesn't allow pictures to be posted directly. Imgur works perfectly.

That PSU is definitely questionable.
When you launch HWiNFO64 put a check in 'sensors only'. While it's running start logging (bottom of window). Go game until it crashes to desktop. Upload log file somewhere we can grab it and have a look.
 
Jun 22, 2022
13
0
10
0
Tom's doesn't allow pictures to be posted directly. Imgur works perfectly.

That PSU is definitely questionable.
When you launch HWiNFO64 put a check in 'sensors only'. While it's running start logging (bottom of window). Go game until it crashes to desktop. Upload log file somewhere we can grab it and have a look.
This is a very old PC. I only got this AMD GPU a little over a year. Back when I had a Nvidia GPU I never had this issue, my Nvidia GPU worked with no problems till it just became faulty (it was very old).

I'll do what you said and I'll report back to you later.
 
Jun 22, 2022
13
0
10
0
What OS version are you running? What was the old GPU model?
Did you use DDU to uninstall the old NVIDIA GPU?
Link for the log: Filebin | 3vg8ijxslygnxz4t
I'm using Windows 10, latest version. Old GPU was a Nvidia GT 710 iirc.

Also yeah, it has been long years since then and I installed several new copies of the OS since. My new SSD has a fresh windows 10 install and it has the same problem.
 
Is it crashing after just 2 minutes of gaming?
Weird that your CPU goes to 100% on all cores right when the GPU usage goes to 0. Could be a side affect of the crash though (if it's crashing at that time).

Download and run Prime95 stress test. Be sure to turn off all the AVX settings and run a blended test. Does it crash within a couple minutes?
 
Jun 22, 2022
13
0
10
0
Is it crashing after just 2 minutes of gaming?
Weird that your CPU goes to 100% on all cores right when the GPU usage goes to 0. Could be a side affect of the crash though (if it's crashing at that time).

Download and run Prime95 stress test. Be sure to turn off all the AVX settings and run a blended test. Does it crash within a couple minutes?
It goes to 100% sometimes or close to it when I'm playing Age of Empires 2 (Which is what I tried when I got driver timeout). Also no, I was just really ''lucky'' that it happened to crash right as I launched it. Sometimes I can play for hours nonstop without nothing happening. It's just random.

When I open Torture Test with Prime95 the (weaker torture test) which has the AVX options are greyed out, so I'll run a blend of the normal test. I heard this Test can take hours so I might let it happen while I go to sleep.

Edit: I let it run for 30 mins real quick and it didn't find any errors. I'll let it run for some hours when it's time I go to sleep, just to make sure.
 
Last edited:
It goes to 100% sometimes or close to it when I'm playing Age of Empires 2 (Which is what I tried when I got driver timeout). Also no, I was just really ''lucky'' that it happened to crash right as I launched it. Sometimes I can play for hours nonstop without nothing happening. It's just random.

When I open Torture Test with Prime95 the (weaker torture test) which has the AVX options are greyed out, so I'll run a blend of the normal test. I heard this Test can take hours so I might let it happen while I go to sleep.

Edit: I let it run for 30 mins real quick and it didn't find any errors. I'll let it run for some hours when it's time I go to sleep, just to make sure.
If it ran for 30 mins you're CPU is probably okay. Next let's test memory. Download and create a Memtest86 USB. Boot from the USB and let it go through at least one full pass to test the memory. After that download and run Unigine's Heaven benchmark. Let it loop for at least 30 mins to test the GPU.

Your PSU is still suspect but can't say for sure. Do you have another system (a friend's, maybe?) that you could test the GPU in? The GPU could also be faulty. Run the above tests to narrow it down more.
 
Jun 22, 2022
13
0
10
0
If it ran for 30 mins you're CPU is probably okay. Next let's test memory. Download and create a Memtest86 USB. Boot from the USB and let it go through at least one full pass to test the memory. After that download and run Unigine's Heaven benchmark. Let it loop for at least 30 mins to test the GPU.

Your PSU is still suspect but can't say for sure. Do you have another system (a friend's, maybe?) that you could test the GPU in? The GPU could also be faulty. Run the above tests to narrow it down more.
I will do as you say.

On a side note, after my last reply with the crash, I played for 7 hours straight without any Driver timeout, so it's very random when it happens, which makes it hard to test on someone else computer since I can't really take their computer for such a long time (I've went a whole day without a single timeout before), especially when they use it regularly.

Edit: I tested both. No problems with memtest, but I got a driver timeout message with heaven benchmark but the benchmark was not stopped and just continued as if nothing happened, but that would be no different from getting driver timeout while playing.
 
Last edited:
Jun 22, 2022
13
0
10
0
If it ran for 30 mins you're CPU is probably okay. Next let's test memory. Download and create a Memtest86 USB. Boot from the USB and let it go through at least one full pass to test the memory. After that download and run Unigine's Heaven benchmark. Let it loop for at least 30 mins to test the GPU.

Your PSU is still suspect but can't say for sure. Do you have another system (a friend's, maybe?) that you could test the GPU in? The GPU could also be faulty. Run the above tests to narrow it down more.
So. I got my cousin to help me and I got the GPU running Heaven Benchmark there for roughly 4 hours with no driver timeout. So I guess something else in my computer is causing this. I really have no idea specially since you told me to run tests for my CPU and RAM, which didn't detect any problem.
 
Last edited:
The randomness of the issue (sometimes crashing after a couple mins sometimes lasting 7+ hours) is a VERY hard issue to diagnose.

Did you perform the memory test?
Maybe try another full Windows wipe and reload but, before putting all your normal programs on it or making adjustments, just install the latest drivers and the game you have the most issue with and test that. Maybe you've got a misbehaving app?
You should also test with a better power supply.

Oh, how's your power grid? Do you get brownouts? Are you running your computer from a UPS?
 
Last edited:
Jun 22, 2022
13
0
10
0
The randomness of the issue (sometimes crashing after a couple mins sometimes lasting 7+ hours) is a VERY hard issue to diagnose.

Did you perform the memory test?
Maybe try another full Windows wipe and reload but, before putting all your normal programs on it or making adjustments, just install the latest drivers and the game you have the most issue with and test that. Maybe you've got a misbehaving app?
You should also test with a better power supply.

Oh, how's your power grid? Do you get brownouts? Are you running your computer from a UPS?
Yes, I did perform the memtest, full pass with no problem. I've tried multiple windows wipes with barely anything installed and I still get the problem.

I don't get any brownouts and testing another power supply will be a bit hard, also not running computer on a UPS.
 
Last edited:
This is going to be tough to pin down without a separate GPU or PSU.

One more thing to note, the GPU bandwidth is being starved by that motherboard. The GPU supports PCIe 3.0 while it looks like the motherboard is only running at 1.0. Depending on the material I read that board should be capable of up to PCIe 2.0. Can you check out BIOS to see if you can change your PCIe x16 slot to the highest version supported?
 
Jun 22, 2022
13
0
10
0
This is going to be tough to pin down without a separate GPU or PSU.

One more thing to note, the GPU bandwidth is being starved by that motherboard. The GPU supports PCIe 3.0 while it looks like the motherboard is only running at 1.0. Depending on the material I read that board should be capable of up to PCIe 2.0. Can you check out BIOS to see if you can change your PCIe x16 slot to the highest version supported?
The only PCI option I was able to find in the BIOS was to choose primary graphic adapter. It had Onboard, PCI Express and PCI. It was on PCI and I just changed to PCi Express right now. Altho this probably isn't what you meant. I didn't find any thing similar to what you suggested me to look for and change.
 

Cyberat_88

Distinguished
Apr 9, 2011
1,774
34
19,890
48
The hunt in Gaming is for the Most Compatible driver, not the newest driver.
Try the Radeon 19.6.1 for the RX 500 series.
But do upgrade that PSU for a name brand with warranty min. 650W.
 

ASK THE COMMUNITY

TRENDING THREADS