Announcement

Collapse
No announcement yet.

15-Way AMD/NVIDIA Graphics Card Comparison For 4K Linux Gaming

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

  • 15-Way AMD/NVIDIA Graphics Card Comparison For 4K Linux Gaming

    Phoronix: 15-Way AMD/NVIDIA Graphics Card Comparison For 4K Linux Gaming

    Being in the middle of working on Linux reviews for the NVIDIA GeForce GTX 980 Ti and AMD Radeon R9 Fury, there's been a lot of fresh graphics processor benchmarks running this week at Phoronix. As the first of these updated large Linux comparisons on the very latest public drivers, here is a 15-way NVIDIA GeForce and AMD Radeon graphics card comparison when running various Linux games with a 4K resolution.

    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
    Just as usual, the long bars is nvidia, the short bars AMD. Except for Civ:Beyond Earth, courtesy of rendering errors.

    Comment


    • #3
      Originally posted by bug77 View Post
      Just as usual, the long bars is nvidia, the short bars AMD. Except for Civ:Beyond Earth, courtesy of rendering errors.
      I'm really curious to see how the R9 Fury will perform in the mix.
      Michael Larabel
      https://www.michaellarabel.com/

      Comment


      • #4
        Is this a joke? You go cheap on AMD and expensive on Nvidia. Bravo.

        Comment


        • #5
          Originally posted by Marc Driftmeyer View Post
          Is this a joke? You go cheap on AMD and expensive on Nvidia. Bravo.
          Huh? Did you read the article? Specifically:
          The 15 graphics cards used for today's comparison were the latest AMD/NVIDIA graphics cards within my possession that were available for testing and from recent generations (the newer AMD/NVIDIA GPU cards not part of the comparison are busy part of the daily performance tracking test lab setup at LinuxBenchmarking.com). The line-up included:..There are a lot more NVIDIA GPUs tested than AMD simply because NVIDIA Corp sends out review samples practically on every different graphics card while AMD hasn't been supplying any Linux review samples in the past few years; all my GCN GPUs except for the R7 370 (supplied by MSI) were bought by Phoronix Media. There are the R9 Fury results coming next week as my latest graphics card purchase, but if you wish to see more AMD Linux graphics card reviews on Phoronix, please consider subscribing to Phoronix Premium or making a PayPal tip.

          If you want to see more AMD GPU tests, then consider donating/subscribing/etc. The GPUs I usually buy for AMD tests are mid-to-high-end as can't always afford buying the top end AMD cards while this time I "splurged" for the R9 Fury due to interest in Fiji/HBM.
          Michael Larabel
          https://www.michaellarabel.com/

          Comment


          • #6
            Originally posted by Michael View Post

            Huh? Did you read the article? Specifically:
            The 15 graphics cards used for today's comparison were the latest AMD/NVIDIA graphics cards within my possession that were available for testing and from recent generations (the newer AMD/NVIDIA GPU cards not part of the comparison are busy part of the daily performance tracking test lab setup at LinuxBenchmarking.com). The line-up included:..There are a lot more NVIDIA GPUs tested than AMD simply because NVIDIA Corp sends out review samples practically on every different graphics card while AMD hasn't been supplying any Linux review samples in the past few years; all my GCN GPUs except for the R7 370 (supplied by MSI) were bought by Phoronix Media. There are the R9 Fury results coming next week as my latest graphics card purchase, but if you wish to see more AMD Linux graphics card reviews on Phoronix, please consider subscribing to Phoronix Premium or making a PayPal tip.

            If you want to see more AMD GPU tests, then consider donating/subscribing/etc. The GPUs I usually buy for AMD tests are mid-to-high-end as can't always afford buying the top end AMD cards while this time I "splurged" for the R9 Fury due to interest in Fiji/HBM.
            The only other AMD GCN cards I own are the R7 260X, R9 270X, and HD 7850,which are in systems in the daily benchmarking farm, but wouldn't add much value anyways considering the R7 370 and HD 7950 are covered.
            Michael Larabel
            https://www.michaellarabel.com/

            Comment


            • #7
              Bioshock Infinite? This game is not good for benchmark for now.
              Why?
              Simple.
              On Radeon with Catalyst all eON game have disable ARB_texture_storage. This OpenGL extesion is broken on catalyst and VP disable it in Witcher2, bioshock etc.

              Looks like new Catalyst 15.7 finally fix it. Look at this topic with devs VP answer: https://github.com/virtual-programmi...ment-120726143

              But if we want test eON games we need still make some trick, because games still have disable this extension. So we need doing some trick with change name binary.

              Michael, could you look at this? Gives this any better performance on Catalyst 15.7? Look at solution on github.

              Comment


              • #8
                Originally posted by xpris View Post
                Bioshock Infinite? This game is not good for benchmark for now.
                Why?
                Simple.
                On Radeon with Catalyst all eON game have disable ARB_texture_storage. This OpenGL extesion is broken on catalyst and VP disable it in Witcher2, bioshock etc.

                Looks like new Catalyst 15.7 finally fix it. Look at this topic with devs VP answer: https://github.com/virtual-programmi...ment-120726143

                But if we want test eON games we need still make some trick, because games still have disable this extension. So we need doing some trick with change name binary.

                Michael, could you look at this? Gives this any better performance on Catalyst 15.7? Look at solution on github.
                Wasn't aware of that Catalyst bug, but it's still interesting test though for comparing between the different NVIDIA (or AMD) graphics cards.
                Michael Larabel
                https://www.michaellarabel.com/

                Comment


                • #9
                  Yeah try '--eon_disable_catalyst_workarounds' switch to see what you get.

                  For the matter of benchmark consistency, not that i care about those wrapped to driver games
                  Last edited by dungeon; 22 July 2015, 05:18 PM.

                  Comment


                  • #10
                    Originally posted by article
                    [...] the AMD Catalyst driver is still in need of more optimizations as outlined a few days ago in How To Make CS:GO Run Much Faster On AMD Catalyst For Linux.
                    But was this trick used in the benchmark itself or not?

                    Comment

                    Working...
                    X