Announcement

Collapse
No announcement yet.

Mesa Threaded OpenGL Dispatch Finally Landing, Big Perf Win For Some Games

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

  • Mesa Threaded OpenGL Dispatch Finally Landing, Big Perf Win For Some Games

    Phoronix: Mesa Threaded OpenGL Dispatch Finally Landing, Big Perf Win For Some Games

    Four years ago Intel developers were working on a threaded OpenGL dispatch mechanism for Mesa, but it never ended up being merged. Now, prolific Mesa contributor and AMD developer Marek Olšák is looking to merge this code and clean it up...

    http://www.phoronix.com/scan.php?pag...hread-Dispatch

  • #2
    This could help with The Talos Principle as well, right?

    @Michael: I got the message when posting here:
    {"nodeId":929798}
    Last edited by GraysonPeddie; 02-05-2017, 11:01 PM. Reason: :

    Comment


    • #3
      Are plans for it to be enabled in the future?

      Comment


      • #4
        I wonder if this would also solve the Rocket League performance problems? (https://bugs.freedesktop.org/show_bug.cgi?id=97879)

        Comment


        • #5
          Interesting tension here between CPU and GPU rendering: CPUs are multi-threaded, GPUs are not.

          Comment


          • #6
            Originally posted by GraysonPeddie View Post
            This could help with The Talos Principle as well, right?
            It won't in its current shape. But actually it can replace Talos and Serious Sam 3 multithreaded rendering. For testing, you can disable Talos's threaded rendering with cvar "ren_bMultiThreadedRendering=0" and than enable this driver's one to get similar/same results

            It is not smart enough yet (i think currently might help only to some games who does not do threading GL at all)... this is only Marek's call to merging it as it improve perf in some cases, but it can be improved in future to eventually speed up more cases.

            Originally posted by jbennett View Post
            I wonder if this would also solve the Rocket League performance problems? (https://bugs.freedesktop.org/show_bug.cgi?id=97879)
            No, that is something else bug It is not related neither to shader cache nor this threaded GL features.
            Last edited by dungeon; 02-06-2017, 12:18 AM.

            Comment


            • #7
              I hope this helps with ETS2!

              Comment


              • #8
                Originally posted by ldo17 View Post
                Interesting tension here between CPU and GPU rendering: CPUs are multi-threaded, GPUs are not.
                No, in fact completely the opposite in the most extreme way !!!

                This change is to do with the single threaded nature of the GL **API** it has nothing to do with hardware CPU vs. GPU.

                Comment


                • #9
                  Originally posted by funfunctor View Post
                  This change is to do with the single threaded nature of the GL **API** it has nothing to do with hardware CPU vs. GPU.
                  Well, in so far as described in the article, in theory this technique can be used with any API, single or multi-threaded. Or at any level of an application's internal or external call stack. However in the case of Vulkan, it would be less likely a meaningful way to split up the work onto multiple threads, since other options exist.

                  Usually, I think, you would dispatch work on a higher level. To do it at the external API level is a way that can be implemented by a library below the application level, transparent to the application. In a sense, it is a substitute for multithreading within the application, and may in some cases work very well, and in others not so well (involving unnecessary copying of data, for example).

                  Makes me wonder if it could be combined with GLVND to optionally benefit any/all OpenGL drivers on Linux.
                  Last edited by indepe; 02-06-2017, 12:24 AM.

                  Comment


                  • #10
                    Originally posted by indepe View Post

                    Well, in so far as described in the article, in theory this technique can be used with any API.
                    It's already used by nine I believe

                    Comment

                    Working...
                    X