[SOLVED] XFX RX 580 crashing on Doom Eternal and several other games

Status
Not open for further replies.

TarikS

Commendable
Apr 12, 2020
19
0
1,510
Hey everyone, today I kept experiencing an annoying and a pain in the a$$ problem. While playing Doom Eternal, I encountered a similar issue I had playing games before but not severe as this one. So what happens is that my monitor goes dark and enters rest mode, then the gpu would full blast the fans and then the PC will freeze while the fans are blasting. This happened to me once or twice before but in Doom it happens every 10-20 minutes for some odd reason. I have no idea why this keeps happening. Does anybody know a fix to this issue, could it be a problem with something else that crashes the GPU?
PC Specs:
Aorus B450
Ryzen 5 1600AF
XFX 580 4GB
Kingston Fury 2x8GB
Be Quiet L7-530W

Edit: also to mention the fans are normal speed while playing the game, then when it crashes they go to 100% and I have to manually turn off the PC
 
Last edited:
be-quiet-pure-power-l7-530w-5.jpg

265W max to the GPU.

Are you sure you don't have a faulty stick of RAM?

Have you applied any custom settings to the GPU? Undervolting/overclocking/power limits/ etc?
 
be-quiet-pure-power-l7-530w-5.jpg

265W max to the GPU.

Are you sure you don't have a faulty stick of RAM?

Have you applied any custom settings to the GPU? Undervolting/overclocking/power limits/ etc?
No adjustements to any of the components, only xmp enabled, i bought the ram new 5/6 months ago
 
Ran a torture test to the cpu, and a 60 sec stress test to the gpu (not sure if enough) so im gonna try running furmark. The cpu is stable and no problems whatsoever.
 
New doesn't mean it's not faulty.

GPU driver updates available?

Monitor CPU & GPU temps.
Updated the drivers, cpu temps normal, 85C on prime 95 100% usage, gpu temps on 83 at 100% usage and it crashed again with furmark, deffo something going on with the gpu, it also now crashed the gpu completely saying
There is a problem with Radeon RX 580 Series device. For more information, search for 'graphics device driver error code 22'. Had this problem once a while back and it fixed by itself
 
Fixed the gpu, it disabled itself in device manager and just needed to be reenabled and the pc restarted, no idea what the hell is it causing it to crash, the temps are good, everything seems fine but it keeps crashing on me. I think I am gonna try to underclock it.
 
Fixed the gpu, it disabled itself in device manager and just needed to be reenabled and the pc restarted, no idea what the hell is it causing it to crash, the temps are good, everything seems fine but it keeps crashing on me. I think I am gonna try to underclock it.
Check your PCIe power wires. See they're not bent, broken inside or smell like burnt. Pay special attention to the plug and all the pins. I recently had a client with the Gigabyte RX 580 with the same exact issue: when there was no load (surfing the net, watching videos) the GPU would work normally, but under any load it would simply crash.

If all the wires are fine, I'd try the GPU with another PSU to rule it out as a possible casue.
 
Check your PCIe power wires. See they're not bent, broken inside or smell like burnt. Pay special attention to the plug and all the pins. I recently had a client with the Gigabyte RX 580 with the same exact issue: when there was no load (surfing the net, watching videos) the GPU would work normally, but under any load it would simply crash.

If all the wires are fine, I'd try the GPU with another PSU to rule it out as a possible casue.
Will check that, just did the underclocking, it did absolutely nothing, 0 difference.
 
Also to mention the core speed is 1366 mhz and vram speed 1750, only lowered the core speed by -50 and vram stayed the same.
Pull the frequency down to 1100MHz at least for testing purposes.

Ran a torture test to the cpu, and a 60 sec stress test to the gpu (not sure if enough) so im gonna try running furmark. The cpu is stable and no problems whatsoever.
So you ran a 60 second test to diagnose an issue that happens every 10-20minutes?

If all the wires are fine, I'd try the GPU with another PSU to rule it out as a possible casue.
This is kindof where I was going with my first post. PSU is definitely on my list of suspects.

Also, have you actually ran any RAM stability/fault tests? If so, what program(s), and for how long?
 
Pull the frequency down to 1100MHz at least for testing purposes.


So you ran a 60 second test to diagnose an issue that happens every 10-20minutes?


This is kindof where I was going with my first post. PSU is definitely on my list of suspects.

Also, have you actually ran any RAM stability/fault tests? If so, what program(s), and for how long?
So I switched the gpu power connector to the secondary one, reseated the ram, and after running furmark the core speed naturally dropped 100+ mhz because of the secondary connector and it crashed after 3-4 minutes of running. I have no idea what happened to the gpu. Keep in mind that I had this GPU in 2 computers, with the same PSU and this is after i bought the new cpu, mobo and ram that issues like this doom example happened. I have no clue what is up, currently dont have another PSU, could it perhaps be with something else, maybe I should try a different GPU and see how its acting up.
 
So good news guys, I finally fixed the problem. What I did was:
  1. Reset the shader cache with the Radeon Adrenaline program
  2. Set the shader setting to low (from ultra)
The game works flawlessly now, even got more fps due to shader cache reset. (Could have been messed up by upgrading from Win10 to Win11)

I think that reseting the shader cache will fix 90% of FPS and crashing problems. Thank you all for helping me through this head hurting adventure. Peace ✌️
 
Status
Not open for further replies.