Announcement

Collapse
No announcement yet.

What Shall We Benchmark Next? Let Us Know!

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

  • #61
    'Linux and other cluster benchmarks. Like MPI, (i dont see possible performance variance of openMP sense they use OS threads anyway) HPLinpack could be a start. you could benchmark networking equipment like that. or use Intel MPI Benchmark. You could make your own mpi benchmark tests to compare MPI implementation. Otherwise, ya compare bsd's new filesystems like Hammer and ZFS and such. Compare ZFS under bsd and [open]solaris and compare those to BTRFS and TuxFS. if you can Compare redhat's GFS and Orical's ClusterFS.

    Comment


    • #62
      Please test the validity of the statement that compiling your kernel for your architecture does improve performance. My whole family can see the difference. But the majority can't seem to notice it or they aren't admitting to it.

      Also test with High-Mem disabled for systems under 1 Gigabyte of memory.
      And again test it with it enabled.

      Comment


      • #63
        Selinux

        I don't think your test suite includes tools to pin test a SELINUX installation. But that would be a nice test to see if all the hype lives up to an actual reputable authority on well testing.

        Comment


        • #64
          As mobile Linux based platforms continue to rise sharply in popularity, I think it would be really useful to benchmark mobile OS-es like Maemo (MeeGo), Android, WebOS and maybe others.

          Comment


          • #65
            Extensive Opensource ATI vs closed source and ATI vs Nvidia benchmarks on a monthly/bi-monthly basis .

            Comment


            • #66
              Tst ARM vs. x86

              I would be highly interested in reading tests showing the performance and energy consumption of ARM vs. x86 ( netbook / tablet segment). I've never seen a real comparison showing the ARM architecture could reach the performance of the Atom CPU's, just the marketing is saying they consume less. There are a lot of new attendees such as Qualcomm Snapdragon, Nvidia Tegra (2?) or Marwell 610. The performance of the graphical processing would be interesting as well (not video decoding).

              Comment


              • #67
                I'd love to see a Debian testing VS Ubuntu comparison

                Comment


                • #68
                  Another suggestion:

                  Video playback comparing the opensource vs the commercial codec performances.

                  Fluendo just released their Codec Pack version 11 and it has support for both VDPAU and VAAPI. I'd like to see what advantage I'd get with my ATI card since I have no player VAAPI support and I don't like compiling/patching.

                  Comment


                  • #69
                    I know you already track the linux kernel but I'd like to see the FreeBSD kernel aswell and benchmarks between the two.

                    Comment


                    • #70
                      Definitely it'd have to be compilers. I'd be very interested in seeing the different versions and the differences in compile and runtime of both C and C++ for some non-trivial size programs.

                      Comment


                      • #71
                        I'd suggest to benchmark the 4500MHD with the latest drivers and compare linux vs xp results.
                        Many people have this vga on their laptop, and has it can't play the newest games anyway, it would be interesting to see which platform (linux or xp) will offer the better performance, and to evaluate how much effort in Intel putting to develop linux drivers.

                        Comment


                        • #72
                          Nvidia vs ATI proprietary drivers on both Linux and Windows.

                          Comment


                          • #73
                            I would strongly suggest an ongoing tracing of ClangBSD vs FreeBSD (that is, LLVM/Clang vs GCC) with an as identical code base as possible. Hot off the presses: ClangBSD is self-hosting!

                            It would be really interesting to see which parts that would generate better binaries from the respective compiler.

                            Comment


                            • #74
                              Looks like a Wine tracker shall be coming in the near future now that we have the backing of a few key Wine contributors.

                              http://www.winehq.org/pipermail/wine...il/083230.html
                              Michael Larabel
                              http://www.michaellarabel.com/

                              Comment


                              • #75
                                Originally posted by Michael View Post
                                Looks like a Wine tracker shall be coming in the near future now that we have the backing of a few key Wine contributors.

                                http://www.winehq.org/pipermail/wine...il/083230.html
                                Now that is really awesome Michael! Regressions are surely one of Wine's biggest problems.

                                Comment

                                Working...
                                X