Weird freezes on a 5700XT sapphire pulse

My pc started getting some weird freezes when playing games lately and i’m not sure how to debug. I can’t pinpoint when it started to happen either because due to work i didn’t play any games for a few months, but here’s what happens:

  • Whenever i play Final Fantasy XIV on Linux, after a few minutes the PC freezes. Those freezes behaves like when you’re using too much memory, video stops, audio gets frozen on loop and then keyboard stops responding. Except memory and cpu usage never go over 30% on any game
  • If i open Pathfinder: Wrath of the righteous, at some random point (even on menus, before the game loads), everything will freeze just like on FFXIV. If i get lucky and load the early game, it will freeze during initial cinematic. FPS is normal when it does load.
  • If i open Rome: Total war 2 i can browse menus without any hiccups, but the moment o load a game it freezes. I haven’t been able to load a match, as soon as it starts loading it freezes.

It’s also worth noting that some games works on Linux without any issues, i haven’t tried many, but so far Dark Souls 1 and Pathfinder Kingmaker both run fine. If you want me to try one specific kind of game (using a certain engine or something) just tell me what kind and i’ll find one in my library.

This is all on linux, here’s the kicker though, Rome: Total war 2 works flawlessly on Windows, but Pathfinder freezes on windows too. This is what got me confused.

Here are the things i’ve tried:

  • Different distros
  • All kernels from 517 until 514
  • Doing a factory reset on GPU settings
  • disabling DX10, DX11, and vulkan
  • different ram frequencies
  • Proton versions until the games don’t even open anymore (at around 5.0)

As for system and debug info, ill provide anything anyone needs. At this point I’m justcurious as to what the hell is going on so if you have any ideas I’ll try them.

Steam system info: Computer Information: Manufacturer: Micro-Star International Co., Ltd. - Pastebin.com
Proton log when attempting to load a total war map (couldn’t use pastebin because it’s just too much text): Computer Information: Manufacturer: Micro-Star International Co., Ltd. - Pastebin.com

Is it just those particular games, or are those listed just random examples?

If it is other games doing it, maybe the GPU is overheating when the GPU usage spikes as it’s loading into the main game.

Those are just the ones i’ve tested, but i can test more.

As for overheating, it doesn’t really have time to overheat. As soon as the game starts to load, the fans are berely spinning, everything freezes.

That’s what I’m getting at. Usually when a game goes from menu to actual game load, there’s a usage spike. If during that spike for some reason the temp sensor that controls the fans doesn’t detect the temperature change, the GPU temp shoots up but the fans don’t do anything other than basic spinning.

Just a thought.

hmm, makes sense, do you know how could i test that under linux? on windows i can use radeon’s overlay to monitor these things, but on linux i couldn’t find any monitoring tool with overlay.

Sorry, no clue. I don’t use Linux … at the moment. I’m just looking at it from a hardware view since I don’t know any testing methodology with Linux.

1 Like

I believe in the USA that the 5700xt from Sapphire has a 3 year warranty. The 5700xt came out in 2019, so it is very likely that your warranty is still good. I suggest you check when you purchased your card and contact Sapphire for help. SAPPHIRE::::Customer Service Center

1 Like

I’m in Brazil though, and i use it for work, since it’s only breaking on games i couldn’t justify sending it over to the US.

But yeah, that would definitely be ideal.

1 Like

Still, it could not hurt to contact Sapphire and see if they will replace it under warranty.

If you can dual boot, I’d troubleshoot the card from the Windows side of things first. There are a lot more diagnostic tools available and as you mentioned you can get frame rate and probably temperature overlays in-game. Maybe just start with something simple like Heaven benchmark and HWMonitor to see if there are thermal issues with the GPU?

1 Like

I just tried it on windows and now i’m completely lost as to what could possibly be happening. My suspicion was fans not spinning and thus high temperature, neither are happening.

When booting wrath of the righteous for the first time, it gives you a screen to configure screen brightness, on windows i can’t even get past the screen. Using AMD’s overlay i could see that the fans were at 815rpm and temperature was 55º. Physically, everything seems normal.

At this point i’m just going to wait until next gen to play games again when i get a new pc. It really doesn’t seem worth the money to buy a 6800xt or something.

PSU or thermals
Get a multimeter out.

Sorry to ask more specifically.

While the games done go over 30% memory, does something else and then the games pushes it over.

Or could it be a driver, you don’t get any Blue Screen errors?

Not that I understand the arcane, but does event viewer show any errors that could point.

My money is actually on your system ram being just stable enough to ead you to believe it’s not that, they recently found a new niche way ram can cause instability and are including a test for it in the memtest version

So run a memtest

Could be the card but could also not be the card

I can put it under heavy loads in other scenarios. Lost ark is a game that does not bug out, but raises temperature and general usage, i can play it for 12 hours without any hiccups.

Not that i noticed while playing with resource overlay. All resources are below expected levels.

That was i was betting on too, i disabled all overclock features, ram is running at stock frequency, nothing changed. I’ll run a memtest later today, didn’t know about those scenarios.

I didn’t say anything about load. If you have a multimeter and can’t be bothered to use it, stop asking for advice.

I’m not sure what corner of 4chan you came from but a reply is not a personal attack, i do not have a multimeter nor any indication that the issue is PSU or thermals, as either would occur under either heavy loads or random moments. These freezes happen under very specific and controlled circumstances.

You understand PSUs can be problematic when not under heavy load right?

1 Like

Yes, i do. Do you agree that problematic PSUs behave unpredictably? Either they have a threshold in which they can’t operate anymore or they fail during normal usage at seemingly random intervals. None of which is applicable for my situation.

My experience begs to differ. Brand trust experience/ opinions +other rambling (low effort video, I'm not redoing it, fuck off) - YouTube Seemed to happen more often under load, but that wasn’t THE reason.