Announcement

Collapse
No announcement yet.

Radeon vs. NVIDIA With Windows 10 & Ubuntu Linux

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

  • Radeon vs. NVIDIA With Windows 10 & Ubuntu Linux

    Phoronix: Radeon vs. NVIDIA With Windows 10 & Ubuntu Linux

    A few days back we published some Windows 10 vs. Linux gaming benchmarks with two Radeon graphics cards. For putting those numbers into better perspective, here are the results now when adding in two competing NVIDIA graphics cards on both operating systems.

    http://www.phoronix.com/vr.php?view=25598

  • #2
    Basically OGL finds D3D11 between 2k and 4k ultra when the overhead is no more a block. Vulkan can do that even on 1080p ultra, that is why i consider VK9 and VK11 to be a win.

    Comment


    • #3
      Because gaming benchmarks depends on games, I would use only Tomb Raider 2013 that works equally good in all platforms. If phoronix test games where the win virus hoover is the winner it makes Linux less attractive platform for newbies who focus only to fps and fight against viruses, installs and remove software one by one and so on.

      Comment


      • #4
        The 4.15 kernel has a performance penalty for the Radeon cards. Despite this they still show a fight against Nvidia, especially the RX580 defeating the GTX1060 in most tests.

        Michael, in the future, when you do a next batch of Linux vs Windows benchmarks, please consider using VP game ports too. They matured very much and today apear to be more competitive than Feral ports.

        Comment


        • #5
          Xonotic on Linux was faster with the Radeon graphics cards while under Windows the NVIDIA driver was much faster.
          That's only true for the Vega card, though. It also seems that the windows driver is hitting a bottleneck in this case.

          Thank you for these benchmarks

          debianxfce, the whole point of a benchmark isn't to make a platform look good, but to compare them. That's why the PTS exists, to have reproducible results. That way, we can see where the bottlenecks are, and where there is still room for improvement. It also needs to be representative of what the user might get on a platform/card combo.

          Comment


          • #6
            Originally posted by M@GOid View Post
            Michael, in the future, when you do a next batch of Linux vs Windows benchmarks, please consider using VP game ports too. They matured very much and today apear to be more competitive than Feral ports.
            Please provide some examples. All VP games than I've played did not updated for years. Also glthread in several eON ports doesn't work for me despite being officially enabled (Witcher 2 and Spec Ops: The Line are among them).
            Last edited by puleglot; 12-03-2017, 10:50 AM.

            Comment


            • #7
              Thank you Michael. Wish I was in position to give you some tips

              Comment


              • #8
                Good tests

                At simple seek GTX 1060 have better min fps (in various test have better max fps but min fps is more important) compared RX 580 without forget RX 580 higher tdp*

                *RX 580 is more or less RX 480 overclocked and stock RX 480 have higher tdp compared GTX 1060

                Comment


                • #9
                  Originally posted by puleglot View Post
                  Please provide some examples. All VP games than I've played did not updated for years. Also glthread in several eON ports doesn't work for me despite being officially enabled (Witcher 2 and Spec Ops: The Line are among them).
                  Arma 3 has an "experimental" VP port that is being updated regularly.
                  Performance is on par with Windows on my system (Ryzen 1700X, 32GiB DDR4-2400, R9 Fury, Ubuntu 17.10).
                  Recommend downloading the "Yet another Arma benchmark" aka. YAAB from Steam workshop for benchmarking.
                  Just like on Windows, the game is very CPU-dependent, but doesn't care about core count. It's all about single threaded CPU performance.
                  Last edited by Brisse; 12-03-2017, 11:03 AM.

                  Comment


                  • #10
                    Michael
                    Have you ever tried benchmarking games on Clear Linux? I wonder if those optimizations have any impact on gaming.

                    Comment

                    Working...
                    X