My computer with an rx580 has been having some serious black-screening problems for a while now. I’ll be playing a video game for a little bit, and the monitor will randomly go black, then display a message a couple seconds later saying something about it going into low power mode. When this happens the fans in the computer will speed up and become noticeably louder. The only way for me to get out of this when it happens is to hold the power button down on the case for a few seconds and start it back up. I have tried almost everything I think I can possibly try.
I have an:
1. rx580 4Gb
3. 32 gb of physical ram
4. msi z390 gaming plus mobo
5. ssd and hdd for storage
6. not the best psu but i’m not getting a new one for a while.
things i’ve tried:
1. updating windows
2. updating drivers (obviously)
3. using older driver versions
4. installing different drivers with and without the amd software
5. limiting the performance of my gpu through the software
6. increasing fan speed (temp not a problem)
7. checked power plan and other windows settings
8. probably some other things i can’t remember
1. i’ve had this problem for so long, i’m not sure when it started. Since it has started though, I have gotten a new processor (i5-???? —> i9-9900), a few new mobos (one got fried and one had the cmos battery slot broken on arrival), new sticks of ram (3x8 —> 2x16), new ssd (500gb m.2), a new cpu cooler, and a new monitor.
2. Gpu temp reaches a max of about 65 celsius under high stress, so i don’t think it’s overheating.
3. cpu reaches upwards of 85 under high stress but after doing stress tests geared towards the processor, i have found that it overheating does not cause the crash.
4. I also ran a gpu stress a few times and found that it didn’t cause the crash.
5. I vaguely remember running some sort of memory test and it crashed, but i can’t recall for sure.
I tried to provide as many details as I can, but if anyone knows anything or needs more info please let me know!
I just messed around with stress tests a bit and noticed that it is something to do with the gpu. It runs just fine for as long as I want with no other programs open (excluding hwinfo and voicemeeter banana). I can open google chrome and tab in between the stress test and chrome just fine, but not discord. I open discord and then try to tab from the stress test to discord and it black screens almost instantly. The temperature of my graphics card reaches about 68 celsius and doesn’t climb any higher. Normally it just black screens randomly while playing a game. In GTA it often happens right when a cinematic begins in a heist for example. It even happens in Overwatch 2 when I limit the fps to 60 (I can get about 110 fps when unlimited).
Common things to try:
- disable hardware acceleration from Chrome, Edge, Firefox, Steam and other software
- disable MPO https://community.amd.com/t5/drivers-software/amd-you-can-do-better-and-more-stable-drivers/m-p/5510...
- if you have Apex Legends installed (or perhaps EA in general), uninstall it, DDU drivers and reinstall after uninstalling Apex Legends
- use only one monitor
- some get help from disabling HDR
- stop alt-tabbing while gaming
Try doing the following:
Press Win + R to open the Run application.
Type regedit into the text field and press Enter. You will need administrator rights to make changes to the Windows registry.
In the registry, navigate to HKEY_CURRENT_USER > Software > Microsoft > Avalon.Graphics using the navigation bar on the left.
Here, right-click on an empty space in the right pane and select New > DWORD (32-bit) Value.
Name the entry DisableHWAcceleration.
Now double click on the record you just created and enter 1 in the Value data field.
Click the OK button.
Exit Registry Editor and restart your computer.
If after this method you are still getting a black screen error, then the problem is probably hardware, so you need to enable hardware acceleration again by changing the data value in the DisableHWAcceleration entry to 0.
"The disable hardware acceleration option enables you to turn off hardware acceleration for debugging and test purposes."
Ok. One more option to do. Had to check what it is about.
This is not an option, but in 95% of cases the main cause of black screens and video driver crashes. A lot of time was spent to determine that this is the reason for the unstable operation of AMD video cards.
I thought heat was the culprit because I pegged 85c in Zero Horizon Dawn. Looked at the fan curve and adjusted it, it kept the heat down to 75c but still did the blank screen intermittently. That's when things got more frustrating and I tried various things until one lone YouTuber with a **bleep**ty video (and music) pinpointed the issue to the boost clock. Soon as I adjusted that to base like he did and boom, issue all but disappeared so far.
It possible I may have gotten a bad chip in terms of turbo boast but the thing never produced artifacts which would have pinpointed to the GPU being pushed too hard. There's something else tied to it, since it's not a heat issue once the fan curve was adjusted. Not drivers, as I tried both 2.61 and 2.81 & 2.82. it's literally narrowed down to just the turbo boost in my situation.
Try disabling MPO and forcing off vsync globally. Not sure if it will help you, but it seems that hw acceleration is conflicting for you at some point.
You can also try to disable hw acceleration in Discord.