Announcement

Collapse
No announcement yet.

Ubuntu 13.10 32-bit vs. 64-bit Performance

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

  • #31
    Originally posted by synaptix View Post
    How is there no way to optimize it?
    use result to prove i am wrong. I would suggest phoronix to test x86-32bit router vs x86-64bit router (download both images from www.pfsense.org)

    Comment


    • #32
      Originally posted by synaptix View Post
      Really, is that why I can get more FPS on games running Mint 15 and I get better system responsiveness and my shutdowns don't hang like it does on Ubuntu?

      Oh and is that why all my packages and software matches that of current up to date version on Ubuntu as well?
      1) Different WM, that is why, for both gaming and system responsiveness. Any WM not using compositing will have very similar performance in games.

      2) The packages are the same, but the releases are delayed compared to Ubuntu.

      On Xubuntu 13.04, I don't have any hang-on-shutdown problems, but generally I wait a bit (a few weeks) before reinstalling so that the showstopper bugs are fixed (and the servers being less loaded).

      That said, Mint does look interesting and has a few good points.

      Comment


      • #33
        Originally posted by Calinou View Post
        1) Different WM, that is why, for both gaming and system responsiveness. Any WM not using compositing will have very similar performance in games.

        2) The packages are the same, but the releases are delayed compared to Ubuntu.

        On Xubuntu 13.04, I don't have any hang-on-shutdown problems, but generally I wait a bit (a few weeks) before reinstalling so that the showstopper bugs are fixed (and the servers being less loaded).

        That said, Mint does look interesting and has a few good points.
        System responsiveness is the Same on MATE for me when using compositing or not it's Gaming is around 1 to 3 FPS Differents in Games i can upload a Video for you if you want

        the delayed is to have a clean install on Linux Mint and not a beta install like on UBUNTU

        Comment


        • #34
          Originally posted by LinuxGamer View Post
          System responsiveness is the Same on MATE for me when using compositing or not it's Gaming is around 1 to 3 FPS Differents in Games i can upload a Video for you if you want
          That's with "real" fullscreen (enjoy your impossibly to Alt+Tab in 90% of games). When playing windowed or with "fake" fullscreen (eg. Source games), you'll definitely notice a difference.

          Comment


          • #35
            Originally posted by Calinou View Post
            That's with "real" fullscreen (enjoy your impossibly to Alt+Tab in 90% of games). When playing windowed or with "fake" fullscreen (eg. Source games), you'll definitely notice a difference.
            i can Alt Tab in any of my games what are you talking about?

            Comment


            • #36
              Originally posted by oliver View Post
              Oh stop it already you guys. Ubuntu is his reference point, it's the most popular distribution and that is just fine. Occasionally there are cross-distro comparissions, which is good enough.

              What does this benchmark try to show us? That x86_64 is a bit ahead of i386, and I bet that's the same across the line. Personally, I'm not bothered one bit by the use of Ubuntu.

              Generally, more important would be (in graphics tests, not the case here anyway) is the various compositers. I notice that gnome-shell/unity3D do hit 3D performance quite a bit, so on tests involving 3D there should be the standard reference system used, and one where for example mate is running, e.g. uncomposited.

              The only thing really missing in this benchmark, is x32, if that's what its called. e.g. x86_64 with 32bit memory pointers. The best of both worlds, was it not?

              So hint for next time Michael, there's 3 options now for x86 testing, i386, x86_64 and x86_32. Though I'll admit, x86_32 is quite harder to test, since I don't think Ubuntu does that yet.
              I agree, Ubuntu is just a choice. I don't see any fanboyism, he needs a distro to test with and he chooses Ubuntu. It would be great to see how Mate, Cinnamon, KDE, Gnome 3, Unity, XFCE, LXDE, etc all stack up for desktop speed in various benchmarks. Not an easy task but the results would be interesting.

              Comment


              • #37
                I freely admit my reasons for remaining with 32-bit are minor. They mostly consist of 32-bit games and emulators that require additional steps to run on 64-bit Linux. The fact is, there is nothing about 64-bit that jumps out as a huge advantage to me in either leisure or serious categories. In fact, most everything I do with my computers, both leisure and serious, can be done just fine on one of my 486 machines. The fastest computer I have is a Pentium 4 (good old single core technology). There's only a couple of games I play on it that won't run great on older machines.

                Comment


                • #38
                  http://www.gamingonlinux.com/article...-and-more.2198

                  Comment


                  • #39
                    Originally posted by Maegashira View Post
                    can we please discuss the failure of the apache benchmarch, which is a big deal for systems running apache ... like ... this page? or any page?
                    There's not any point. The setup of the Apache benchmark Phoronix does is explicitly against what Apache recommends. It's in a configuration that exactly 0 people actually use in the real world. And it seems to fluctuate rather randomly based on kernel version. All of which point to the test being completely useless.

                    Comment


                    • #40
                      So if I'm on a 32 bit Ubuntu 12.04 install, what are the steps to convert to 64 bit?

                      Comment


                      • #41
                        Originally posted by RecceDG View Post
                        So if I'm on a 32 bit Ubuntu 12.04 install, what are the steps to convert to 64 bit?
                        Reinstall from scratch.

                        Well, you can keep your home partition around if it's separate from root if you want to try to keep your user settings.

                        Comment


                        • #42
                          32-PAE vs 6B, but with 8 or 16 GB of memory?

                          Guessing - but it seems that 32-PAE is going to be much slower with 8 or more gigabytes of RAM, doing a large memory-demanding task(s).

                          Most times my PC is multi-tasking: bacFkground Internet watching, various timers and indicators, or background backups, etc ...

                          Other unusual notes on this test: was the operating system Ubuntu, or Xubuntu? Big difference, since disabling Ubuntu's Unity is complex.

                          Also unusual is the use of the latest stable kernel. This is not part of the default distributions. I update all my distros to the latest stable kernel ASAP too. But then my Dell XPS-16 notebook needs the kernel's SSD-friendly benefits, not available with normal distro kernels.

                          Pity some juniors here can't read. As already mentioned, it is really a beginner's test of 32 vs 64 distro. Unknown seemingly to the tester, is that 64-bit is useful with more than 4GB of RAM. 32-PAE was created AAIK, since 64bit op systems had not developed enough for early adopters, until just recently.

                          Some apps use as much memory & cores as much is free. Anyone knows if 32-PAE is slower, with us guessing?
                          Last edited by gregzeng; 08-19-2013, 07:10 AM.

                          Comment


                          • #43
                            Originally posted by bridgman View Post
                            I can think of two possibilities :

                            1. The poster could be talking about colour blindedness -- Red-green is the most common form in humans, and affects males more than females.

                            2. The poster could be a Red Green fan, but neither handsome nor handy.
                            {grin}
                            Neither is Red Green :P

                            Comment


                            • #44
                              Originally posted by synaptix View Post
                              Actually it was master to master copy entire ~/.steam folder to ~/Documents/.steam.

                              I only use slave as reserve backup.

                              Also this was both on Kernel 3.11-rc2 as well. I've always had slow transfers and system unreponsiveness on vanilla Ubuntu, which never happens on Xubuntu or Mint.
                              Out of curiosity, how were you transferring those files?
                              Nautilus ? (G)rsync ? Bash 'cp' command ?

                              I can easily see Nautilus causing problems, especially in Ubuntu.
                              In fact, a couple of bad experiences with Nautilus in Ubuntu was what caused me to brush up on rsync, again, and make using that standard practice for large file transfers, even between disks/partitions on just one computer...

                              Comment

                              Working...
                              X