Announcement

Collapse
No announcement yet.

Ubuntu 21.10 Radeon Gaming With KDE Plasma vs. GNOME Shell + Wayland vs. X.Org

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Ubuntu 21.10 Radeon Gaming With KDE Plasma vs. GNOME Shell + Wayland vs. X.Org

    Phoronix: Ubuntu 21.10 Radeon Gaming With KDE Plasma vs. GNOME Shell + Wayland vs. X.Org

    With last week's release of Ubuntu 21.10, here are some fresh benchmarks looking at the Linux gaming performance on this new release while testing both the GNOME Shell 40 default desktop to that of its KDE Plasma 5.22 based option. Both the X.Org and Wayland sessions for KDE and GNOME were benchmarked for seeing how the Linux gaming performance compares with the Radeon open-source GPU driver stack.

    Phoronix, Linux Hardware Reviews, Linux hardware benchmarks, Linux server benchmarks, Linux benchmarking, Desktop Linux, Linux performance, Open Source graphics, Linux How To, Ubuntu benchmarks, Ubuntu hardware, Phoronix Test Suite

  • #2
    That's great.
    Everything looks the same other than a few bad apples on KDE.

    Comment


    • #3
      I am astonished to see Wayland loosing against Xorg. Shouln't by Wayland more lean, simple and in result, faster, than bloated/olded Xorg?

      Comment


      • #4
        Hitman 2 results look odd to me (at least on Xorg), Proton games should just trigger suspend KWin compositing. Wayland benchmarking might be questionable until dmabuf-hints land: https://invent.kde.org/plasma/kwin/-...82#note_307483
        There's no use of having nice fps when it stutters badly regardless.

        Comment


        • #5
          Originally posted by Ilfirin View Post
          I am astonished to see Wayland loosing against Xorg. Shouln't by Wayland more lean, simple and in result, faster, than bloated/olded Xorg?
          I don't remember games being ported to use Wayland. XWayland adds some Xorg to Wayland

          Comment


          • #6
            So Plasma has equal or better perfomance than gnome shell in almost all tests....in ubuntu, not a friendly kde distro.

            Comment


            • #7
              Originally posted by Ilfirin View Post
              I am astonished to see Wayland loosing against Xorg. Shouln't by Wayland more lean, simple and in result, faster, than bloated/olded Xorg?
              The games run on xwayland, a Xorg layer on wayland. Your statement can be true once games ship and optimize for native wayland.

              Comment


              • #8
                Originally posted by Ilfirin View Post
                I am astonished to see Wayland loosing against Xorg. Shouln't by Wayland more lean, simple and in result, faster, than bloated/olded Xorg?
                High performance 3D rendering applications go out of their way to bypass as much of the display server as possible. As long as everything works as expected, there should be no difference in raw performance between X11 and Wayland. Native Wayland apps might perhaps deliver a bit less stuttering but that's not something that'd show up when measuring just average FPS.

                Comment


                • #9
                  Originally posted by Ilfirin View Post
                  I am astonished to see Wayland loosing against Xorg. Shouln't by Wayland more lean, simple and in result, faster, than bloated/olded Xorg?
                  If it was pure Wayland then there would be a reason to be concerned, but since these games are typically running through Xwayland, the Wayland compositor is basically just added overhead.

                  So really, the takeaway here is Wayland's overhead is astonishingly minimal.

                  Comment


                  • #10
                    Originally posted by Yeayo guy View Post
                    So Plasma has equal or better perfomance than gnome shell in almost all tests....
                    Kwin is architectural a lot more modern. In the end the values here however are "the same" across all the tested desktops and displayservers. It's just to minimal of a difference.

                    Comment

                    Working...
                    X