Announcement

Collapse
No announcement yet.

Dell XPS 13: Windows 10 vs. Linux Distribution Benchmarks

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

  • Dell XPS 13: Windows 10 vs. Linux Distribution Benchmarks

    Phoronix: Dell XPS 13: Windows 10 vs. Linux Distribution Benchmarks

    Recently I have published benchmarks looking at Windows Server and FreeBSD against eight Linux distributions as well as a 9-way Linux desktop OS benchmark comparison while the latest in this string of fresh Linux distribution benchmarks is looking at the Linux laptop performance impact, if any, between these operating systems. Up for this benchmarking dance was Microsoft Windows 10, Windows 10 when running Ubuntu 18.04 via WSL, Ubuntu 18.04 itself, Fedora Workstation 28, openSUSE Tumbleweed, and Clear Linux.

    http://www.phoronix.com/vr.php?view=26585

  • #2
    I won't let it be:

    Originally posted by phoronix View Post
    Ubuntu on WSL under WIndows 10 ended up being
    Originally posted by phoronix View Post
    Windows 10 caried quite a decent punch.
    It makes me sad to see Microsoft's implementation of Linux faster than Torvalds Linux...

    Comment


    • #3
      Originally posted by tildearrow View Post
      It makes me sad to see Microsoft's implementation of Linux faster than Torvalds Linux...
      It's not the case according to the conclusion, Clear Linux is ahead most of the times, which means it's not the implementation but the compiler and the compilation options that make the difference here.

      Comment


      • #4
        Originally posted by spykes View Post
        It's not the case according to the conclusion, Clear Linux is ahead most of the times, which means it's not the implementation but the compiler and the compilation options that make the difference here.
        FYI Clear Linux's optimizations aren't solely due to compiler/compiler-optimizations but much more.
        Michael Larabel
        http://www.michaellarabel.com/

        Comment


        • #5
          I'm running openSUSE Tumbleweed on an old laptop with a Sandy Bridge i3 with a Linux-libre kernel, and I cut your PyBench time nearly in half. Does my ext4 file system account for all that difference?


          Comment


          • #6
            Those iGPU benchmarks don't seem right. When I tested my Skylake notebook on Win10 and Linux, it was marginally faster on Linux.

            That difference is quite large, so I suspect a power management issue? e.g. Windows boosts for longer? Or maybe HuC/GuC isn't set up properly (it isn't used by default on Linux due to causing issues with some hardware, but on mine it gave me a noticeable performance boost for iGPU)

            Comment


            • #7
              @Michael: Would be cool if you could take the relative performance into account in your conclusion. "Clear Linux was the fastest most often with 45% of the time being in first" doesn't tell us by which amount it was better. My impression is, that Windows often outperformed Linux by quite some amount, while when Linux leads, Windows wasn't that much behind. So overall, it seems to me that Windows sadly performed best... -.-

              Comment


              • #8
                Originally posted by Michael View Post
                FYI Clear Linux's optimizations aren't solely due to compiler/compiler-optimizations but much more.
                It's open source though, right? I assume there's some patches for the kernel or something, would be cool to know where. I'm starting to kind of get tired of the bad options that mainstream distros use (e.g. Ubuntu) that slow down the kernel so want to compile a Clear Linux kernel on a "normal" distro.

                Seriously I can't believe Windows 10 performed better (except for the video/gpu related stuff) but I guess it's what you get when performance takes a back seat to "maintainability" these days.

                Comment


                • #9
                  Originally posted by Michael View Post

                  FYI Clear Linux's optimizations aren't solely due to compiler/compiler-optimizations but much more.
                  Can you elaborate on this ?
                  They are using a patched libc? A patched kernel? Both?

                  Comment


                  • #10
                    Originally posted by spykes View Post

                    Can you elaborate on this ?
                    They are using a patched libc? A patched kernel? Both?
                    Yes they have patches against both. I have covered some of their optimizations in various articles, there is also some documentation on ClearLinux.org, and if going to their GitHub page you can see their dozens of patches against many different components.
                    Michael Larabel
                    http://www.michaellarabel.com/

                    Comment

                    Working...
                    X