Announcement

Collapse
No announcement yet.

Benchmarks: 2 BSDs vs. 7 Linux Distributions

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

  • Benchmarks: 2 BSDs vs. 7 Linux Distributions

    Phoronix: Benchmarks: 2 BSDs vs. 7 Linux Distributions

    My latest benchmarking enjoyment has been testing two BSD operating systems against seven Linux distributions on the same Intel Haswell system. Here are those latest benchmark numbers.

    http://www.phoronix.com/vr.php?view=23439

  • #2
    And Linux is faster when it comes to most tests, not amazing...

    Comment


    • #3
      Simple disk IO load: FreeBSD > Linux. (Postgres, apache, fileservers)
      CPU math intensive: Linux > FreeBSD. (Any CPU intensive load).

      So what's the conclusion? Linux disk I/O & VFS suck slow bananas and Clang @ FreeBSD produces slow binaries?

      Comment


      • #4
        Would be awesome to include some Linux distros with btrfs (FreeBSD having ZFS and DFBSD having HAMMERFS).

        Comment


        • #5
          Network performance tests would also be very cool. So we can finally put to rest the big question bugging all us, "is BSD's network stack really as superior to Linux as Veteran Unix Admins believe?" and focus on more important matters again afterwards.

          Originally posted by milkylainen View Post
          So what's the conclusion?
          That apart from a few tests BSD is only marginally worse, so any claim that BSD has better CPU utilization because of better code/schedulers/whatever is busted unless it's stuff you only see in 64-core beasts so it's not relevant here, although it's still not a so big gap to be embarassing.

          I'm not seeing any graph about simple disk IO, these are all CPU tests.

          Comment


          • #6
            awesome to see Manjaro here again since its the best alround-desktop linux out there

            Comment


            • #7
              My favorite distro Manjaro did OK, but Ubuntu did wey better, NOT GOOD!

              Comment


              • #8
                But why test BETA builds?
                I'm not sure how FreeBSD betas are build now, but 5.x-9.x betas were built without optimization enabled.

                Comment


                • #9
                  Originally posted by blacknova View Post
                  But why test BETA builds?
                  I'm not sure how FreeBSD betas are build now, but 5.x-9.x betas were built without optimization enabled.
                  You can see my latest 10.3 vs. 11 Beta benchmarks here where the performance came down close to the same, which is why just one was shown for this comparison: http://www.phoronix.com/scan.php?pag...nchmarks&num=1
                  Michael Larabel
                  http://www.michaellarabel.com/

                  Comment


                  • #10
                    No Debian Testing? You're gonna trigger someone.

                    Comment

                    Working...
                    X