882
Steam data reveals PC gamers shifting from Windows to Linux
(www.tweaktown.com)
you are viewing a single comment's thread
view the rest of the comments
view the rest of the comments
this post was submitted on 04 Aug 2025
882 points (98.2% liked)
PC Gaming
15460 readers
449 users here now
For PC gaming news and discussion. PCGamingWiki
Rules:
- Be Respectful.
- No Spam or Porn.
- No Advertising.
- No Memes.
- No Tech Support.
- No questions about buying/building computers.
- No game suggestions, friend requests, surveys, or begging.
- No Let's Plays, streams, highlight reels/montages, random videos or shorts.
- No off-topic posts/comments, within reason.
- Use the original source, no clickbait titles, no duplicates. (Submissions should be from the original source if possible, unless from paywalled or non-english sources. If the title is clickbait or lacks context you may lightly edit the title.)
founded 3 years ago
MODERATORS
Anyone know if CP2077 runs better on Linux than Windows?
By much? With HDR?
Sorry for the drive by comment, but this is like the one game my 3090 can’t quite handle to my satisfaction. I've thoroughly disabled the thing from rendering in Linux and don’t want to undo all that… But if I could get like another 10% over Windows, that would be incredible. Even 5% would be awesome.
As I promised, my own Cyberpunk testing of Windows Vs Linux on mostly the same hardware (they are on different SSDs, but I don't think that'll have a drastic impact).
TLDR: Windows framerates seem inconsistent, it's first benchmark I ran (the first Ultra without DLSS) was way faster with no explanation. Aside from that and Ray Tracing: Overdrive, Linux seems to win, and by a large degree (+28 FPS average on the Low preset seems ridiculous).
I don't think these results are broadly applicable to more machines. You probably won't get +28 FPS by switching to Linux.
My best guess is that the performance difference may have a lot to do with different power/thermal targets, or that Windows was doing a lot in the background (it was running an update, but I didn't expect a huge impact).
I'm guessing that on most hardware the performance difference will be pretty small.
Hardware: ROG Zephyrus G15 GA503QR Laptop Ryzen 9 5900HS, 16 GiB DDR4 RTX 3070 Laptop GPU 2560x1440 screen, up to 165 Hz
All benchmarks: plugged into OEM power supply. I held the laptop vertically so there were no restrictions to its airflow.
Game: Cyberpunk 2077 V2.3 with Phantom Liberty DLC, fullscreen 2560x1440. Values are given as Min / Average / Max FPS displayed by the game's built in benchmark.
Linux (Bazzite 42): NVIDIA driver 575.64.05 Samsung 980 Pro 2TB SSD Performance power profile
Low Preset ( no upscaling): 57.49 / 68.42 / 83.86 FPS
Ultra Preset(no upscaling): 32.91 / 39.27 / 49.71 FPS
Ultra (DLSS Transformer model, Auto): 41.11 / 48.70 / 61.30 FPS
Ray Tracing: Low Preset (DLSS transformer model, Auto): 44.12 / 51.70 / 61.63 FPS
Ray Tracing: Ultra Preset (DLSS transformer model, Auto): 29.24 / 34.26 / 39.81 FPS
Ray Tracing: Overdrive Preset (DLSS transformer model, Auto): 15.03 / 17.71 / 20.45 FPS
Windows (Windows 11 Home 23H2): GeForce Game Ready Driver 580.88 SK Hynix HFM001TD3JX013N SSD "Turbo" power profile (in ASUS Armoury Crate)
Low Preset (no upscaling): 35.68 / 40.68 / 45.17 FPS
Ultra Preset(no upscaling): 40.53 / 52.88 / 65 FPS
Ultra Preset (no upscaling, Round 2): 29.68 / 35.63 / 39.94 FPS
Ultra (DLSS Transformer model, Auto): 36.71 / 47.20 / 55.32 FPS
Ray Tracing: Low Preset (DLSS transformer model, Auto): 28.55 / 32.41 / 35.85 FPS
Ray Tracing: Ultra Preset (DLSS transformer model, Auto): 22.23 / 27.25 / 30.86 FPS
Ray Tracing: Overdrive Preset (DLSS transformer model, Auto): 17.74 / 19.96 / 22.64 FPS