Announcement

Collapse
No announcement yet.

The May 2012 Open-Source Radeon Graphics Showdown

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

  • The May 2012 Open-Source Radeon Graphics Showdown

    Phoronix: The May 2012 Open-Source Radeon Graphics Showdown

    Up for publishing today is a multi-card multi-driver comparison spanning several generations of discrete ATI/AMD Radeon graphics cards looking at the Radeon OpenGL performance as found by default in Ubuntu 12.04, as found when updating to the latest "Git" development code, when tweaking the latest development code for maximum performance, and finally when using the proprietary AMD Catalyst Linux display driver.

    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
    Thanks for including the tweaked versions Michael. Would be great if you could write a few lines about power profiles. From the test overview table it looks like all cards have been running at full speed. Then the more than 8 times slower performance looks almost like a bug.

    Comment


    • #3
      No screenshots of the Prey corruption?

      Comment


      • #4
        What happened to 6450 here? I don't understand it at all! How can catalyst always reach 2.5x factor performance, but on some titles fall to 70% of it? I have a feeling catalyst is through and through based on hacks, cut-offs and similar. Really strange.

        Btw, has somebody actually have any luck with video acceleration on open-source amd drivers?


        Btw, radeon 64 ddr (r100) is completely broken in kernel 3.0 (ubuntu 11.04).

        Comment


        • #5
          Acceleration

          On the newer oss drivers that support the vdpau state tracker, vdpauinfo reports that acceleration is in place and works. However, it does not yet support h264 so it's usefulness is limited, but is rapidly improving.

          Obviously, xvba or vaapi acceleration is not present with these drivers.

          Comment


          • #6
            How do I turn on 2D color tiling? I saw Phoronix articles a few weeks back on this, and decided to try it. The most logical-looking thing I could add to xorg.conf was bounced out as an invalid option. I've already got PCI-e 2 enabled.

            Comment


            • #7
              Originally posted by crazycheese View Post
              What happened to 6450 here? I don't understand it at all! How can catalyst always reach 2.5x factor performance, but on some titles fall to 70% of it? I have a feeling catalyst is through and through based on hacks, cut-offs and similar. Really strange.

              Btw, has somebody actually have any luck with video acceleration on open-source amd drivers?


              Btw, radeon 64 ddr (r100) is completely broken in kernel 3.0 (ubuntu 11.04).
              I personally would be lucky if the open source drivers even worked at all. I use a 5750 (unlocked as a 5770) with KDE 4 under 64 bit Arch Linux, and I can't even see the desktop. I know it isn't a compositing issue because if I run LXDE I get the same problem. I'm getting an ati-dri update now so hopefully that works, but I've been booting from an 8400GS for the past month.

              Comment


              • #8
                These numbers actually look quite encouraging for me, especially since the largest resolution I play at is 1280x1024.

                Comment


                • #9
                  Why Radeon Git Tweaked tested without swapbufferswait off?
                  Lightsmark 2008 shows 130 fps on hd6770
                  1205193-BY-GOG20120520

                  Comment


                  • #10
                    Originally posted by Pontostroy View Post
                    Why Radeon Git Tweaked tested without swapbufferswait off?
                    Lightsmark 2008 shows 130 fps on hd6770
                    1205193-BY-GOG20120520
                    Interesting. The test log shows: Option "SwapbuffersWait" "0"

                    Comment

                    Working...
                    X