Announcement

Collapse
No announcement yet.

A Variety Of OpenGL/OpenCL NVIDIA 367.27 vs. AMD Linux 4.7 + Mesa Git Benchmarks

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

  • A Variety Of OpenGL/OpenCL NVIDIA 367.27 vs. AMD Linux 4.7 + Mesa Git Benchmarks

    Phoronix: A Variety Of OpenGL/OpenCL NVIDIA 367.27 vs. AMD Linux 4.7 + Mesa Git Benchmarks

    For your viewing pleasure this afternoon are some fresh NVIDIA GeForce GTX 900/1000 benchmarks with the 367.27 display driver compared to various Radeon GCN GPUs using a patched Linux 4.7 kernel and Mesa 12.1-dev Git as of this past weekend...

    http://www.phoronix.com/scan.php?pag...Open-Pre-RX480

  • #2
    Looks a bit extra skewed towards Nvidia when the Nv cards are really expensive and the AMD cards are cheap/old ones (except Fury of course).

    Comment


    • #3
      Originally posted by ernstp View Post
      Looks a bit extra skewed towards Nvidia when the Nv cards are really expensive and the AMD cards are cheap/old ones (except Fury of course).
      It's not skewed in any intentional manner. It's a matter of the cards I have: NVIDIA generally sends out all their hardware where as in this comparison the RX 480 is the only one that was sent out by AMD and the other cards I used I had purchased.
      Michael Larabel
      http://www.michaellarabel.com/

      Comment


      • #4
        Originally posted by ernstp View Post
        Looks a bit extra skewed towards Nvidia when the Nv cards are really expensive and the AMD cards are cheap/old ones (except Fury of course).
        Considering the only "new" line of cards that AMD has released since 2013 are the Fury cards, and that Michael reviewed the Fury, it's not exactly his fault.

        Comment


        • #5
          As we see opensource driver OpenCL fully working, to the matter that results are missing

          I complained to why F1 2015 will be benched when it perform very worse then original game everywhere and now we even benchmark what does not work

          So what is next that does not work to benchmark? Nouveau maybe

          Comment


          • #6
            Originally posted by Michael View Post
            It's not skewed in any intentional manner. It's a matter of the cards I have: NVIDIA generally sends out all their hardware where as in this comparison the RX 480 is the only one that was sent out by AMD and the other cards I used I had purchased.
            Yep, I think that one is on us... trying to get it fixed (hopefully it is fixed now because Michael has an RX 480). Things would look better with a 390X and a Fury X in there.
            Last edited by bridgman; 06-27-2016, 05:31 PM.

            Comment


            • #7
              I'm puzzled: AMD cards perform well under Dota2 / Unigine Heaven 4.0 but rather subpar on other games.
              Is it because developers have easier access to Dota2 / Unigine as they're free? Or am I missing something?

              Comment


              • #8
                Originally posted by gurv View Post
                Is it because developers have easier access to Dota2 / Unigine as they're free? Or am I missing something?
                Better written?

                Comment


                • #9
                  There are a few factors, but the main ones seem to be (a) the extent to which the app was written on and for NVidia hardware (big problem a few years ago, much less so today) rather than being written for portability, and (b) how long the app has been running on the open source drivers so that devs have had time to optimize.

                  Remember that these numbers are open source AMD vs closed source NVidia and that closed source AMD is still to come.

                  Comment


                  • #10
                    Originally posted by gurv View Post
                    I'm puzzled: AMD cards perform well under Dota2 / Unigine Heaven 4.0 but rather subpar on other games.
                    Is it because developers have easier access to Dota2 / Unigine as they're free? Or am I missing something?
                    This is opensource AMD driver vs NVIDIA proprietary driver... Let say like on Windows, if you comparing generic driver for one vendor to the full blown driver for the other vendor

                    Comment

                    Working...
                    X