Announcement

Collapse
No announcement yet.

Intel Sandy Bridge Looks Good On Mesa 9.1

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

  • Intel Sandy Bridge Looks Good On Mesa 9.1

    Phoronix: Intel Sandy Bridge Looks Good On Mesa 9.1

    Last week I delivered benchmarks showing Mesa 9.1 delivers faster Intel OpenGL graphics. The benchmarks in that article were carried out on an Intel Core i7 "Ivy Bridge" system with HD 4000 graphics while since then there have been many requests to have similar tests done on a previous-generation Sandy Bridge system. As a result of those requests, here are benchmarks of an Intel Core i5 "Sandy Bridge" processor with Intel HD 3000 graphics as the Mesa 9.1 performance is compared to the earlier 9.0.2, 8.0.5, and 7.11.2 branches.

    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
    i dont know about you, but my professor would rip of the skin from me while fully concious for comparing things with different settings... cant you get rid of the AA options? or have both, to track AA improvements if made?

    Comment


    • #3
      Really, "sandy bridge looks good" ?

      Originally posted by phoronix View Post
      Phoronix: Intel Sandy Bridge Looks Good On Mesa 9.1
      Really, "sandy bridge looks good" ?
      The numbers are good but the screen remind me every second there is something very wrong,
      I'm not really sure theses numbers have any real values when the tearing problem doesn't appear to be solved.

      Comment


      • #4
        Originally posted by rafirafi View Post
        Really, "sandy bridge looks good" ?
        The numbers are good but the screen remind me every second there is something very wrong,
        I'm not really sure theses numbers have any real values when the tearing problem doesn't appear to be solved.
        Tearing and performance are not the same thing. You can solve one without solving second.

        And tearing is probably way more complex, than mild performance optimization.

        Comment


        • #5
          Originally posted by jakubo View Post
          i dont know about you, but my professor jasa seo would rip of the skin from me while fully concious for comparing things with different settings... cant you get rid of the AA options? or have both, to track AA improvements if made?
          you could not son. try it more hard

          Comment


          • #6
            Nexuiz is slower since Mesa 8.0.5 due to multi-sample anti-aliasing (MSAA) having been implemented for Intel Sandy Bridge by Mesa 9.0.
            I read this in every article about Intel and wonder: can you not turn off MSAA to get an accurate performance comparison?

            Comment


            • #7
              Originally posted by jakubo View Post
              i dont know about you, but my agen would rip of the skin from me while fully concious for comparing things with different settings... cant you get rid of the AA options? or have both, to track AA improvements if made?
              just follow this topic

              Comment

              Working...
              X