Announcement

Collapse
No announcement yet.

GCC 8.0 vs. LLVM Clang 6.0 On AMD EPYC

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

  • GCC 8.0 vs. LLVM Clang 6.0 On AMD EPYC

    Phoronix: GCC 8.0 vs. LLVM Clang 6.0 On AMD EPYC

    At the beginning of January I posted some early LLVM Clang 6.0 benchmarks on AMD EPYC while in this article is comparing the tentative Clang 6.0 performance to that of the in-development GCC 8.0. Both compilers are now into their feature freeze and this testing looked at the performance of generated binaries both for generic x86_64 as well as being tuned for AMD's Zen "znver1" microarchitecture.

    http://www.phoronix.com/vr.php?view=25847

  • #2
    I see from GCC comments there is still some work to be done:
    ;; Currently blocking all decoders for vector path instructions as
    ;; they are dispatched separetely as microcode sequence.
    ;; Fix me: Need to revisit this.
    ;; Fix me: Need to revisit this later to simulate fast path double behavior.

    Comment


    • #3
      Try benching a FLAC to MP3.

      Comment


      • #4
        It would be great to see a comparison between Epyc and Threadripper.

        Would that be possible?!

        Comment


        • #5
          What does those flags under individual benchmark means. I.e.
          1. (CC) gcc options: -O3 -march=znver1 -march=native

          Note that at GCC side generic was also retuned to be more Zen friendly and there are still few patches to come.

          Comment


          • #6
            Originally posted by sabriah View Post
            It would be great to see a comparison between Epyc and Threadripper.

            Would that be possible?!
            AFAIR they have different memory channel, besides different platform, so one would have like 32GB quad-channel RAM and the other 96GB hexa/octa-channel

            Comment


            • #7
              Originally posted by sabriah View Post
              It would be great to see a comparison between Epyc and Threadripper.

              Would that be possible?!
              I believe that in an older article Michael said AMD didn't want Epyc and Threadripper benched against each other.

              Comment


              • #8
                Originally posted by sabriah View Post
                It would be great to see a comparison between Epyc and Threadripper.

                Would that be possible?!
                Possible yes, but Michael posted in another Epyc thread that AMD does not want this.

                Originally posted by Michael View Post
                AMD requested I not run any side-by-side tests of Threadripper and EPYC due to intended for different markets :/ But those knowing how to use OpenBenchmarking.org can likely generate a similar comparison with a few clicks.

                Comment


                • #9
                  Originally posted by vsteel View Post

                  I believe that in an older article Michael said AMD didn't want Epyc and Threadripper benched against each other.
                  This makes me assume the results would be interesting...

                  Comment


                  • #10
                    Originally posted by LinuxID10T View Post

                    This makes me assume the results would be interesting...
                    I am guessing that Threadripper would win but they really are somewhat different markets. I don't know if the test suite would really show the strength of a high end server CPU where it is designed for somewhat different software loads. The test suite seems geared more to desktop software and less of virtual machine farms or web server with hundreds of hits a second where the Epyc would do better.

                    Comment

                    Working...
                    X