Announcement

Collapse
No announcement yet.

AMDGPU-PRO Beta 2 vs. Mesa 11.3 + Linux 4.6: Very Competitive For Linux Gamers

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

  • AMDGPU-PRO Beta 2 vs. Mesa 11.3 + Linux 4.6: Very Competitive For Linux Gamers

    Phoronix: AMDGPU-PRO Beta 2 vs. Mesa 11.3 + Linux 4.6: Very Competitive For Linux Gamers

    Following last week's AMDGPU-PRO 16.20.3 "Beta 2" driver release of AMD's new hybrid driver stack for Linux that makes use of the AMDGPU open-source kernel DRM driver with the closed-source OpenGL driver derived from Catalyst / Radeon Software, I set out to do a fresh open vs. closed-source driver comparison. For the Radeon R9 285, R9 290, and R9 Fury, I compared the performance of this new AMDGPU-PRO driver against Mesa 11.3-devel Git and Linux 4.6 for the latest open-source driver stack.

    http://www.phoronix.com/vr.php?view=23213

  • #2
    i have a question . why not just do away with the proprietary bits? it is because opening the stuff up means risking some IP assets ? Also wouldn't it be easier/convenient/time and money saving to have one unified development effort ?

    Comment


    • #3
      Nice test. Thanks Michael . Look like new AMDGPU-PRO working fine

      Any hope for test AMDGPU-PRO vs old fglrx/catalyst 15.12? You can Install it on Ubuntu 14.04/15.10 and benchmark both. We should see if new driver gives us progress in closed source line.

      Comment


      • #4
        Is it feasible to bisect to find the Bioshock/Dirt regression?
        That would be very good to find before next Mesa release - and a scoop for Phoronix :-)

        Comment


        • #5
          Originally posted by Veto View Post
          Is it feasible to bisect to find the Bioshock/Dirt regression?
          That would be very good to find before next Mesa release - and a scoop for Phoronix :-)
          If I was motivated enough with extra time on my main graphics benchmarking system... As first I would also need to check if it was the Radeon/AMDGPU DRM that introduced it, don't remember if it was in Mesa or the kernel as it was an odd regression.
          Michael Larabel
          https://www.michaellarabel.com/

          Comment


          • #6
            Originally posted by Michael View Post

            If I was motivated enough with extra time on my main graphics benchmarking system... As first I would also need to check if it was the Radeon/AMDGPU DRM that introduced it, don't remember if it was in Mesa or the kernel as it was an odd regression.
            This is the exact kind of things which will earn you quite a few tips once the regression get discovered.
            ## VGA ##
            AMD: X1950XTX, HD3870, HD5870
            Intel: GMA45, HD3000 (Core i5 2500K)

            Comment


            • #7
              Originally posted by darkbasic View Post
              This is the exact kind of things which will earn you quite a few tips once the regression get discovered.
              In the past it's been rather mixed when finding regressions if people really live up to tipping, etc, or not.
              Michael Larabel
              https://www.michaellarabel.com/

              Comment


              • #8
                i've tried amdgpu-pro on a laptop with apu carrizo (i know that carrizo is not official supported, but i would try), the temperature and the performance are a lot worse than AMDGPU opensource.
                i hope that in future when AMDGPU-PRO will reach its final version will improve performance an temperature also on mobile APU and mobile dGPU.

                the performance with the tested GPU are amazing, so i hope that also other AMDGPU will receive this treatment

                Comment


                • #9
                  Originally posted by Michael View Post

                  If I was motivated enough with extra time on my main graphics benchmarking system... As first I would also need to check if it was the Radeon/AMDGPU DRM that introduced it, don't remember if it was in Mesa or the kernel as it was an odd regression.
                  Have you checked if any of the systems running on linuxbenchmarking spotted the regression?

                  If it is the case, this can save you some time besides serving as awesome use-case for the system

                  Comment


                  • #10
                    Originally posted by Michael View Post

                    If I was motivated enough with extra time on my main graphics benchmarking system... As first I would also need to check if it was the Radeon/AMDGPU DRM that introduced it, don't remember if it was in Mesa or the kernel as it was an odd regression.
                    Can't be AMDGPU as even r600 is affected, though a bit differently. The bug must be at a higher level, which is shared by all these drivers. The issue came in somewhere around the 11.0 or 11.1 Mesa release, the game ran fine with the last versions of 10.x, using GL version overrides.

                    Someone also put it on the RadeonSI bugtracker, so the developers must be aware of this.

                    Comment

                    Working...
                    X