While you're here, please consider supporting GamingOnLinux on:
Reward Tiers:
Patreon. Plain Donations:
PayPal.
This ensures all of our main content remains totally free for everyone! Patreon supporters can also remove all adverts and sponsors! Supporting us helps bring good, fresh content. Without your continued support, we simply could not continue!
You can find even more ways to support us on this dedicated page any time. If you already are, thank you!
Reward Tiers:
This ensures all of our main content remains totally free for everyone! Patreon supporters can also remove all adverts and sponsors! Supporting us helps bring good, fresh content. Without your continued support, we simply could not continue!
You can find even more ways to support us on this dedicated page any time. If you already are, thank you!
Login / Register
- Nexus Mods retire their in-development cross-platform app to focus back on Vortex
- Canonical call for testing their Steam gaming Snap for Arm Linux
- Windows compatibility layer Wine 11 arrives bringing masses of improvements to Linux
- European Commission gathering feedback on the importance of open source
- GOG plan to look a bit closer at Linux through 2026
- > See more over 30 days here
- Weekend Players' Club 2026-01-16
- simplyseven - Welcome back to the GamingOnLinux Forum
- simplyseven - A New Game Screenshots Thread
- JohnLambrechts - Will you buy the new Steam Machine?
- mr-victory - Game recommendation?
- JSVRamirez - See more posts
How to setup OpenMW for modern Morrowind on Linux / SteamOS and Steam Deck
How to install Hollow Knight: Silksong mods on Linux, SteamOS and Steam Deck
Also, at least with Proton, the first time you launch the game, shaders need to be compiled (whatever that means :whistle: ), and if you start playing immediately, it can cause stuttering.
I don't think the game is bypassing anything. I guess it's just stressing the GPU in a strange pattern.
gamemoderun wine whatever.exeAlso, if you have an "overclocked edition" of your GPU, maybe you can try setting frequency and voltage back to "factory" settings (look for your GPU model on AMD site, the specs should be there). I know I've had some trouble in the past with some OC edition of GPU (granted, I was using Windows, but bad overclocking is bad on any platform).
55 degrees celsius is abnormally low for running games. How did you measure this? Try MangoHUD overlay to see it in 'realtime' at load inside the game. It also gives you a bunch of other useful metrics, like CPU and GPU load. You can use goverlay to customise it conveniently, or just use the cfg files.
I can only speak for a Nvidia gpu (1070) but after I switched to Linux, I had to set my fan controls manually as the automatic setting basically meant they never went on (I discovered to my shock and horror). Now I have a 3rd party fan curve script that starts up at boot time and it's very aggressive giving me cooler temps yet more noise.
Just some quirks of gaming :)
I myself have one of the better ones on the market, MSI gaming X one, with an aggressive custom fan profile (80%@60C) and I still frequently hit 70-85C with good Noctua case cooling to boot. These things just run hot. I know there is a slight stock OC on it, but from my testing it didn't make much of a difference to undervolt it, and the performance hit is not something I enjoy.
[benchmarks](https://www.guru3d.com/articles-pages/nvidia-geforce-gtx-1070-review,9.html#:~:text=First%20up%2C%20IDLE%20(desktop),40%20Degrees%20C%20is%20nice.)
I do not have any custom cooling on my GPU (Asus Strix something, don't have the exact reference at hand), but my case is a (old) Thermaltake, with plenty of space and quite a few cooling fans. No manual overclock, but the GPU might be slightly overclocked fro manufacturer, and the cooling (I checked for you :wink: ) is done by 3 fans.
Have you checked your gpu loads at the same time? Ideally you should check temps exactly when the load is high, with an overlay like MangoHUD. If you alt tab then the load goes down, or if the game barely uses any gpu, then the temps are kinda meaningless. I would try it with either MangoHUD after observing for a couple minutes of an intensive gameplay, where the GPU load has been high consistently high for a few minutes (basically most non-indie games for the past few years or a benchmark like Unigine, 3dMark), or simply with 'watch -n 2 nvidia-smi' or anything like that that shows it under current load quickly (in a second monitor or something).
I'm not asking these things because it's a case of 'well FOR ME, things are different' and I hate being shown I'm doing something wrong with my rig. In fact, that would be nice, so that things can improve. I'm saying this because the benchmarks are all reporting above 70C @stock (I linked one), which is exactly what I'm seeing too. 55C or lower under load seems extremely, suspiciously low.
Same thing with OP - 55C is extremely low/impossible for a RX 570 according to reviews, which means we cannot rule out thermal throttling if the temp measurements are wrong:
https://www.gamersnexus.net/hwreviews/2884-gigabyte-rx-570-aorus-review-vs-470-580-gtx1060/page-3#:~:text=Starting%20with%20frequency%20versus%20temperature,76C%20under%20a%20torture%20scenario.
edit: weird url stuff, why can't I get it working
Last edited by rojimboo on 7 Mar 2021 at 9:47 am UTC
To be clear, I don't run benchmarks, I don't monitor my GPU when gaming (only fps counter with Steam overlay), because I usually don't care :wink:
(I used to do that under Windows though, typically when upgrading GPU. In all my past checks, I never really went over 55-60°C, more usually around 45-50-ish °C, depending obviously on the game).
Also, keep in mind where are talking about gaming (specifically Elex), not "torture test scnerios" like they usually do on benchmarks.
Edit: Actually, I can reproduce this if I check the temps in nvidia-settings after I quit the game. It goes like this:
idle = 45C
load at 100% after 5 minutes of Witcher 3 = 72C (checked in-game with MangoHUD and verified by alt-tabbing in nvidia-settings)
idle after quitting game after around 5seconds = 54C
Needless to say, this does not represent gaming temps and is not useful when trying to determine whether there is thermal throttling or investigating fan behaviour.
Last edited by rojimboo on 7 Mar 2021 at 11:48 am UTC