Ubuntu 26.04 Aims To Deliver Better NVIDIA Wayland Performance Atop GNOME

If all goes well the upcoming Ubuntu 26.04 LTS release will further enhance the NVIDIA graphics performance under its default GNOME Wayland session. The improvements might be upstreamed to GNOME 50 in time but otherwise it's looking like Ubuntu 26.04 will carry its own patch(es) for improving the NVIDIA Wayland performance.

For the past two months there has been a Mutter merge request to "unblock NVIDIA performance." As explained there:

"Avoid entering gbm_surface_lock_front_buffer before rendering is finished, because on the Nvidia driver it's a blocking function and would hurt main loop performance. The GBM docs are sufficiently vague that such blocking is allowed and so might be expected in bespoke GBM implementations.

Using Nvidia-580, this reduces the time spent blocked per frame from milliseconds to microseconds."

Comments on that merge request have acknowledged improvements. The merge request remains open and feedback positive while remains to be seen so far if it will be merged in time for the GNOME 50 release in March. (In case you missed it, GNOME 50 Alpha shipped yesterday.)

Even if this unblocking work for NVIDIA's driver doesn't make it into Mutter 50, it's looking like Ubuntu could end up carrying the patches on their own -- similar to what they have done in the past with carrying extra optimizations and patching in triple buffering support.

Noted today on the merge request is that Ubuntu 26.04 LTS will likely end up carrying this Mutter performance optimization work for NVIDIA:

"All conflicts with main are fixed but that's all I've had time for so far. Unless anything horrible is discovered I expect to start shipping this as a patch in Ubuntu 26.04. But we still hope to improve secondary GPU performance further, and hope to land it in GNOME 50."

Given Ubuntu 26.04 is a Long Term Support release and now Wayland-only for GNOME compared to Ubuntu 24.04 LTS, it's very important to have a good NVIDIA Wayland experience in this release. Hopefully though any and all worthwhile performance optimizations manage to still make it upstream in time for GNOME 50 to avoid distribution-specific carrying of performance improvements.