Announcement

Collapse
No announcement yet.

Ubuntu's Unity Has Room To Improve Performance

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

  • Ubuntu's Unity Has Room To Improve Performance

    Phoronix: Ubuntu's Unity Has Room To Improve Performance

    Following yesterday's news that Ubuntu 12.10 will drop the Unity 2D desktop, I carried out some quick tests comparing the latest state of the Unity desktop with Compiz against the lightweight Unity 2D desktop that's now being removed. To not much surprise, the composited Unity desktop still has some performance shortcomings for OpenGL workloads compared to Unity 2D.

    http://www.phoronix.com/vr.php?view=17765

  • #2
    It will take some time before this works fully ok, I read somewhere that devs themselves expected the work to bleed over to 13.04. So, be in no hurry for this to work the way it is intended.

    Meanwhile I guess many, many Ubuntu-users will switch to Xubuntu.

    Comment


    • #3
      It would be interesting to know technically why Unity bottlenecks some games and why it doesn't have an effect on others.

      Comment


      • #4
        Originally posted by blackout23 View Post
        It would be interesting to know technically why Unity bottlenecks some games and why it doesn't have an effect on others.
        I would also be interested in how the benchmarks perform without a WM, so that we know what the baseline is. For all we know, Unity 'helps' the framerate (I know, it doesn't, but it illustrates my point).

        F

        Comment


        • #5
          Some games were benchmarked at High resolutions and had unplayable FPS.

          You should test those in lower settings too to find out where the playable resolution is (which is what real users will use).

          Comment


          • #6
            Originally posted by madjr View Post
            Some games were benchmarked at High resolutions and had unplayable FPS.

            You should test those in lower settings too to find out where the playable resolution is (which is what real users will use).
            I have to agree - I can understand 1920x1080 being used as a maximum test, but I somehow doubt that is the resolution that most people use (especially with Intel GMA hardware).

            Comment


            • #7
              More strange results

              Don't have access to unity on my ivybridge test machine, but a comparison is here: http://openbenchmarking.org/result/1...SU-1208163SU36

              Comment


              • #8
                Originally posted by blackout23 View Post
                It would be interesting to know technically why Unity bottlenecks some games and why it doesn't have an effect on others.
                Please be sure you're doing updates on Unity when making this claim - there have been several big updates to it where I've noticed that they listed performance improvements. I had a beef with Unity as well when it dropped my 60+ fps game down to about 40fps unless I was running the game widowed - but they've fixed that since.

                Comment


                • #9
                  The CPU is listed as a quad core, but it's a dual with HT. I thought PTS was supposed to handle that?

                  Comment


                  • #10
                    Originally posted by curaga View Post
                    The CPU is listed as a quad core, but it's a dual with HT. I thought PTS was supposed to handle that?
                    In the system information it reports the number of logical cores, within the more detailed CPU table (not shown in article) is where it shows physicsl vs. logical, instruction set extensions, cache sizes, etc.
                    Michael Larabel
                    http://www.michaellarabel.com/

                    Comment

                    Working...
                    X