Announcement

Collapse
No announcement yet.

Ubuntu 13.04 Desktop Gaming Performance Comparison

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

  • Ubuntu 13.04 Desktop Gaming Performance Comparison

    Phoronix: Ubuntu 13.04 Desktop Gaming Performance Comparison

    In this article are some early benchmark results comparing the OpenGL gaming performance of the Unity, Xfce, Openbox, LXDE, KDE, GNOME Shell, and Enlightenment desktops when running on a recent development snapshot of Ubuntu 13.04. As many earlier benchmarks have shown, the OpenGL frame-rate for Linux games can sway quite greatly depending upon the desktop in use and more specifically the desktop's compositing window manager.

    http://www.phoronix.com/vr.php?view=18435

  • #2
    The bad news: Linux desktops are still unoptimized for gaming. Inconsistent at best, complete random at worst.

    The fun news: KDE is freaking slow. Im already looking forward to the neckbeards suggesting hundreds of configs to change instead of just admitting the facts.

    Comment


    • #3
      Originally posted by funkSTAR View Post
      The bad news: Linux desktops are still unoptimized for gaming. Inconsistent at best, complete random at worst.

      The fun news: KDE is freaking slow. Im already looking forward to the neckbeards suggesting hundreds of configs to change instead of just admitting the facts.
      That's funny, I thought it was one config change called unredirect full screen windows, which is the recommended setting for game, and used to be the default before they switched back to not unredirecting by default.

      Comment


      • #4
        Did XFCE have it's compositing effects enabled? I'm surpised it performs so well compared to OpenBox.

        Comment


        • #5
          Michael, clarify here in the forums: was KDE set to unredirect fullscreen windows or not? If so, great, then obviousl theres a regression. But If not then this entire article can be tossed out because the basic premise is "Window manager doesnt matter."

          Like I really dont bash your articles much but you not setting that option from time to time is mine--and others-- biggest peeve with you because theres a certain line between "Keeping the defaults" and "common sense for benchmarking."

          Comment


          • #6
            Where the tests run in Windowed Mode, or Fullscreen? Compositing DEs like Unity/Shell/KDE have Undirect-Fullscreen rendering modes either on by default, or easily enabled. Obviously, if you compare a windowed game running in a composited WM vs a non-composited one like OpenBox (or Xfwm default), then you're going to get better frame rates (though apparently Wayland addresses this nicely).

            I only skimmed the text, but didn't see anywhere where is listed these details. It would be nice to know for these types of benchmarks.

            Comment


            • #7
              Inquiring minds want to know

              Originally posted by Ericg View Post
              Michael, clarify here in the forums: was KDE set to unredirect fullscreen windows or not? If so, great, then obviousl theres a regression. But If not then this entire article can be tossed out because the basic premise is "Window manager doesnt matter."

              Like I really dont bash your articles much but you not setting that option from time to time is mine--and others-- biggest peeve with you because theres a certain line between "Keeping the defaults" and "common sense for benchmarking."
              Agreed, this really needs to be made clear. I used kde 4.10 for gaming and find it quite snappy - but of course I have window rules that say e.g. "inhibit compositing when playing openarena" etc.

              Comment


              • #8
                i would also like to know if gnome shell was used with mutter or compize?

                Comment


                • #9
                  Nothing interesting here for me as openbox user, only seems like good prove that intel runs great across the board.

                  And also seems like supertuxkart have no problems with fps on intel, like radeon has - in some supertuxkart tests i see there radeon is even slower then on llvmpipe
                  Last edited by dungeon; 01-31-2013, 03:48 AM.

                  Comment


                  • #10
                    so: the less complex the WM the more FPS. What about xdm? If Xfce gives me 10fps more than KDE and xdm gives me 10fps more than Xfce this could be totally worth it for a gaming machine. Please include xdm next time.

                    Comment


                    • #11
                      Originally posted by dungeon View Post
                      And also seems like supertuxkart have no problems with fps on intel, like radeon has - in some supertuxkart tests i see there radeon is even slower then on llvmpipe
                      Which parts? Might want to report those as bugs (to either mesa or stk, depending on the cause).

                      Comment


                      • #12
                        @mark_

                        XDM = X Display Manager, the original login thingy?

                        Comment


                        • #13
                          Originally posted by curaga View Post
                          @mark_

                          XDM = X Display Manager, the original login thingy?
                          yes, the plain standard easy and ugly thingy. That one: http://en.wikipedia.org/wiki/XDM_%28display_manager%29

                          Comment


                          • #14
                            Originally posted by curaga View Post
                            Which parts? Might want to report those as bugs (to either mesa or stk, depending on the cause).
                            This:
                            http://openbenchmarking.org/result/1...PTS-SUPERTUX37

                            Compared with this:

                            http://openbenchmarking.org/result/1...RA-TUONGPV7583

                            And then results with intel:

                            http://openbenchmarking.org/result/1...FO-INTELMESA88



                            Could someone rerun supertuxkart with current radeon on non-unity maybe.
                            Last edited by dungeon; 01-31-2013, 04:19 AM.

                            Comment


                            • #15
                              One thing that I find interresting ist, that gnome-shell is sometimes the slowest and sometimes (one of) the fastest desktop(s) according to these benchmarks. How can that be? Shouldn't there be a uniform slowdown due to composite?

                              Comment

                              Working...
                              X