Announcement

Collapse
No announcement yet.

DiRT Rally 2.0 Linux NVIDIA vs. AMD Radeon 18-Way Steam Play Benchmarks

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

  • DiRT Rally 2.0 Linux NVIDIA vs. AMD Radeon 18-Way Steam Play Benchmarks

    Phoronix: DiRT Rally 2.0 Linux NVIDIA vs. AMD Radeon 18-Way Steam Play Benchmarks

    Dirt Rally 2.0 was released last week by Codemasters as the successor to 2015's Dirt Rally and another title on the EGO Engine. While Dirt Rally saw a native Linux port, Dirt Rally 2.0 hasn't seen any port announced by Feral Interactive (though they are currently porting DiRT 4 to Linux), but what's very exciting is this brand new Windows game runs great under Valve's Steam Play with Proton and DXVK! The experience for Dirt Rally 2.0 is quite great on Linux already thanks to Steam Play and with this being a benchmark-friendly game, here are some NVIDIA GeForce and AMD Radeon benchmarks of this racing game on Linux.

    http://www.phoronix.com/vr.php?view=27661

  • #2
    Good stuff, look forward to the radeon 7 numbers once the drivers are fixed.

    Looks like both windows and Linux have had driver issues with this launch.

    Comment


    • #3
      Nice benchmarks. The mean FPS is a good add on from now on.
      Too bad R7 has issues.

      Comment


      • #4
        Michael What's the problem with Radeon VII?

        Comment


        • #5

          I do not understand why you have so few emotions
          Here a free driver for AMD and proprietary from Nvidia are compared. In the game running through Wine. If I were shown this article in 2014, my eyes were the size of an apple.

          Comment


          • #6
            Ran the same Benchmarks, but seems your profile doesn't apply MSAA 4x if its switched off by user as i got totally different results at first:
            https://openbenchmarking.org/result/...SP-DIRT2TEST20 - CPU bottlenecked with my Ryzen at High settings no MSAA
            GPU usage on high settings was around ~80% and Ultra ~100%

            However when forcing MSAA 4x from the game menus and then running again i'm getting much more similar results:
            https://openbenchmarking.org/result/...SP-DIRT2MSAA60
            Last edited by xpander; 03-04-2019, 09:59 AM.

            Comment


            • #7
              Looks like there is not enough VRAM on VI cards? Or the bottleneck is somewhere else?

              Comment


              • #8
                Originally posted by hakzsam View Post
                Michael What's the problem with Radeon VII?
                With DRM-Next 5.1 and 5.0 Git paired with Mesa Git, the desktop stops responding anywhere from a few seconds after GNOME Shell to launching a game or two. When remoting in, I've been getting frequent VMC page faults. e.g.


                [ 151.208845] amdgpu 0000:03:00.0: [gfxhub] VMC page fault (src_id:0
                ring:158 vmid:3 pasid:32775, for process DirtRally pid 4976 thread
                DirtRally:cs0 pid 4978)
                [ 151.208848] amdgpu 0000:03:00.0: in page starting at address
                0x000080018ebf7000 from 27
                [ 151.208849] amdgpu 0000:03:00.0: VM_L2_PROTECTION_FAULT_STATUS:0x0030113D
                [ 151.208852] amdgpu 0000:03:00.0: [gfxhub] VMC page fault (src_id:0
                ring:158 vmid:3 pasid:32775, for process DirtRally pid 4976 thread
                DirtRally:cs0 pid 4978)
                [ 151.208853] amdgpu 0000:03:00.0: in page starting at address
                0x000080018ebf5000 from 27
                [ 151.208854] amdgpu 0000:03:00.0: VM_L2_PROTECTION_FAULT_STATUS:0x0030113D
                [ 151.208856] amdgpu 0000:03:00.0: [gfxhub] VMC page fault (src_id:0
                ring:158 vmid:3 pasid:32775, for process DirtRally pid 4976 thread
                DirtRally:cs0 pid 4978)
                [ 151.208857] amdgpu 0000:03:00.0: in page starting at address
                0x000080018ebf6000 from 27
                [ 151.208858] amdgpu 0000:03:00.0: VM_L2_PROTECTION_FAULT_STATUS:0x0030113D
                [ 151.208861] amdgpu 0000:03:00.0: [gfxhub] VMC page fault (src_id:0
                ring:158 vmid:3 pasid:32775, for process DirtRally pid 4976 thread
                DirtRally:cs0 pid 4978)
                [ 151.208862] amdgpu 0000:03:00.0: in page starting at address
                0x000080018ebf4000 from 27
                [ 151.208862] amdgpu 0000:03:00.0: VM_L2_PROTECTION_FAULT_STATUS:0x0030113D
                [ 151.208865] amdgpu 0000:03:00.0: [gfxhub] VMC page fault (src_id:0
                ring:158 vmid:3 pasid:32775, for process DirtRally pid 4976 thread
                DirtRally:cs0 pid 4978)
                [ 151.208866] amdgpu 0000:03:00.0: in page starting at address
                0x000080018ebb7000 from 27
                [ 151.208867] amdgpu 0000:03:00.0: VM_L2_PROTECTION_FAULT_STATUS:0x0030113D
                [ 151.208869] amdgpu 0000:03:00.0: [gfxhub] VMC page fault (src_id:0
                ring:158 vmid:3 pasid:32775, for process DirtRally pid 4976 thread
                DirtRally:cs0 pid 4978)
                [ 151.208870] amdgpu 0000:03:00.0: in page starting at address
                0x000080018ebbd000 from 27
                [ 151.208871] amdgpu 0000:03:00.0: VM_L2_PROTECTION_FAULT_STATUS:0x0030113D
                [ 151.208874] amdgpu 0000:03:00.0: [gfxhub] VMC page fault (src_id:0
                ring:158 vmid:3 pasid:32775, for process DirtRally pid 4976 thread
                DirtRally:cs0 pid 4978)
                [ 151.208875] amdgpu 0000:03:00.0: in page starting at address
                0x000080018ebdc000 from 27
                [ 151.208876] amdgpu 0000:03:00.0: VM_L2_PROTECTION_FAULT_STATUS:0x0030113D
                [ 151.208878] amdgpu 0000:03:00.0: [gfxhub] VMC page fault (src_id:0
                ring:158 vmid:3 pasid:32775, for process DirtRally pid 4976 thread
                DirtRally:cs0 pid 4978)
                [ 151.208879] amdgpu 0000:03:00.0: in page starting at address
                0x000080018ebdf000 from 27
                [ 151.208880] amdgpu 0000:03:00.0: VM_L2_PROTECTION_FAULT_STATUS:0x0030113D
                [ 151.208883] amdgpu 0000:03:00.0: [gfxhub] VMC page fault (src_id:0
                ring:158 vmid:3 pasid:32775, for process DirtRally pid 4976 thread
                DirtRally:cs0 pid 4978)
                [ 151.208884] amdgpu 0000:03:00.0: in page starting at address
                0x000080018ebaa000 from 27
                [ 151.208884] amdgpu 0000:03:00.0: VM_L2_PROTECTION_FAULT_STATUS:0x0030113D
                [ 151.208887] amdgpu 0000:03:00.0: [gfxhub] VMC page fault (src_id:0
                ring:158 vmid:3 pasid:32775, for process DirtRally pid 4976 thread
                DirtRally:cs0 pid 4978)
                [ 151.208888] amdgpu 0000:03:00.0: in page starting at address
                0x000080018ebb5000 from 27
                [ 151.208889] amdgpu 0000:03:00.0: VM_L2_PROTECTION_FAULT_STATUS:0x0030113D
                Michael Larabel
                http://www.michaellarabel.com/

                Comment


                • #9
                  Originally posted by Maslou View Post
                  I do not understand why you have so few emotions
                  Here a free driver for AMD and proprietary from Nvidia are compared. In the game running through Wine. If I were shown this article in 2014, my eyes were the size of an apple.
                  Well for one thing, the Dirt series has always been a good performer on AMD. But more importantly, a good journalist should remain unbiased. Michael expresses his disapproval of things where he should, but otherwise I'm glad he remains neutral most of the time. That, as well as being the leader in Linux benchmarks, is why I'm subscribed here.

                  Comment


                  • #10
                    Originally posted by Michael View Post

                    With DRM-Next 5.1 and 5.0 Git paired with Mesa Git, the desktop stops responding anywhere from a few seconds after GNOME Shell to launching a game or two. When remoting in, I've been getting frequent VMC page faults. e.g.
                    Is it a recent regression?

                    Comment

                    Working...
                    X