Replies: 1 comment 1 reply
-
|
Hey @Pyxrs, Performance drops jumping to laptop/mobile GPUs in Bevy can sometimes be tied to the default WGPU backend falling back improperly, or high-DPI scaling rendering way more pixels than necessary. Try explicitly setting the resolution limit or checking if the discrete GPU is actually being used rather than the integrated one! Let me know if you make any progress or if you need another fresh set of eyes on it! |
Beta Was this translation helpful? Give feedback.
1 reply
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
Uh oh!
There was an error while loading. Please reload this page.
-
I recently started using a laptop to work on my 3D bevy game in v0.18.1, where I noticed incredibly bad performance that was not present in other games. I was able to reproduce the issues just by running the 3d scene example; on v0.12.1 and before, I average ~2.3 ms. On v0.13.0 and after, it jumps up to ~16.7 ms average. The performance seems to be flatly about 8 times worse than it should be. In the atmosphere example in v0.18.1 my frame times are over 60ms, with my CPU usage fine but my GPU pinned at 100%. I narrowed down the problematic commit to the same one referenced in discussion #13771, which is 35ac1b1. Their issues were dismissed as a result of their CPU's age, but I appear to have the exact same problem with an AMD chip released in 2025. I also have the same GL crashing issue. I dual booted a KDE Plasma NixOS installation and it has the same problem. Drivers and BIOS are up to date. If it's a wgpu issue, then it exists from v0.19 all the way to v28 which I tested on the main Bevy branch.
Rust version:
cargo 1.95.0-nightly (ce69df6f7 2026-02-12)(I also tested stable)Code I'm running to test the issue:
Should I open an issue with wgpu? Just looking for guidance because I'm a bit out of my depth here :)
Beta Was this translation helpful? Give feedback.
All reactions