Announcement

Collapse
No announcement yet.

Ubuntu vs. Arch Linux On The ASUS ROG Strix G15 / Ryzen 9 5900HX

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

  • #31
    Originally posted by Sholto View Post

    explain what being british has to do with ubuntu it was created by a south african
    Imperialist expansionism.

    Seriously though:

    Canonical address

    If you want to speak to someone right away, you can call or write to Canonical. The London-based office is a good place to start.
    • Canonical Group Limited
    • 5th Floor, Blue Fin Building
    • 110 Southwark Street
    • SE1 0SU
    • London, United Kingdom
    That's the address given on both the Ubuntu and Canonical web sites.

    Comment


    • #32
      Originally posted by skeevy420 View Post
      As an actual Arch Linux user who didn't bring up that argument, "but Arch Linux!" only matters if you know how to configure your system for performance. It helps to know to install stuff like cpupower (for the systemd service) and change the scheduler off of schedutil or ondemand over to performance, to go Perpetually High with patches or to install one of the umpteen AUR kernels, and other things like that.

      And after you do all of that, "but Arch Linux!" still might not mean shit. Just because Arch is more plain and generic by default doesn't mean it'll be faster. Sometimes that not-plain and not-generic stuff other distributions patch in make things faster. Sometimes they don't.

      I use Arch because it gives me full control over my system, because the documentation is really good, because the PKGBUILD format and AUR allow me to quickly and easily try new things in a manner that is better for my system than "make && make install", because when I install a program I'm getting what the developer of the program released and not some Debian maintainers interpretation of that program via 21 patches to set defaults and tweak themes. None my reasons are because Arch is 1337 and faster than the rest. That's what Gentoo is for.

      And as a long-term Phoronix reader, unless there is some crazy bullshit happening most every distribution performs around the same. Pick your distribution based on merit, features, support, community, and capability. Don't pick your distribution because some jackasses on Reddit like to post "btw I use Arch".
      I used to be an Arch user but I now use Manjaro because my time is more valuable now, but honestly the thing that keeps bringing me back to Arch based distros is the PKGBUILD format along with makepkg. As a software developer/engineer, this has been one of the top features of the distro that I sorely miss whenever I have to use anything else.

      Being able to easily patch/hack on upstream packages and have them installed long with the rest of your systems packages is a gods blessing. I have no idea how people can deal with distributions like Ubuntu or Fedora where its so painful to do anything out of the ordinary (i.e. I need a specific version of a compiler or I need to apply a patch to quickly fix a bug).

      Originally posted by Linuxxx View Post
      What most Arch Linux fans fail to realize is that most PC users do NOT want to babysit their OS all day long; they want to setup once and then be good to go!

      That is the reason why the Ubuntu LTS base is so popular in the Linux world (including all derivatives like Linux Mint):
      You get to enjoy an enterprise-grade foundation that is properly tested across a very wide userbase.
      And if it works on your current hardware, then it will keep doing so without random updates breaking it every so often like it always inevitably will on Arch Linux! (Been there, done that...)
      I have had more issues with ubuntu (even LTS) then any other distro to date. You are basically betting your "enterprise stability" is due to pinned outdated package and that hopefully you won't get any problems, because if you do and they are not taken seriously by upstream you are screwed. At least with Arch/arch based distros its easier to apply patches to packages.
      Last edited by mdedetrich; 26 July 2021, 03:08 AM.

      Comment


      • #33
        Guys, I'ma let you finish the arch vs Ubuntu discussion

        but... As you all know perpetually high talks big game (and equally, if not more, backs it up) and now I got the receipts

        I re-did the benchmarks I did prior and this time less noise and more polished

        Tested the stock ubuntu kernel (with and without mitigations), and the CFS vs CacULE (CFS enhanced experimental scheduler) with both idle tickless and full tickless, and 500Hz, 1000Hz, and 2000Hz timers). The CFS and CacULE tests are with my custom supercharged kernel. The CacULE tests are just the CFS scheduler with the CacULE patches on top, nothing different.

        One last thing: these were all run right after each other, no other differences, and verified with htop nothing running in background.

        pts/osbench (misc OS benchmarks that capture overall stuff).

        To run: ./phoronix-test-suite benchmark pts/ctx-clock-1.0.0

        and

        pts/ctx-clock (tests the number of context switches).

        To run: ./phoronix-test-suite benchmark pts/osbench


        Few screenshots for the lazy. You guys are leaving a lot of performance on the table. For what? These changes are all free. And you're on the Phoronix forums. You want the Need for Speed, don't you?

        You can and will do better. I will see to it.









        Comment


        • #34
          Originally posted by mdedetrich View Post
          I have had more issues with ubuntu (even LTS) then any other distro to date. You are basically betting your "enterprise stability" is due to pinned outdated package and that hopefully you won't get any problems, because if you do and they are not taken seriously by upstream you are screwed. At least with Arch/arch based distros its easier to apply patches to packages.
          People using Ubuntu generally don't want to apply patches to packages. I'm with Linus, I just want my distro to work without having to mess with it. Ubuntu gives me that, and still allows me to install a different kernel or newer mesa if I desire it. I've never had a problem with Ubuntu, although I tend to stay with the latest releases or dev releases since I don't like to be too far behind.

          Comment


          • #35
            Interesting. systemd OS vs systemd OS, and the winner is ---- no one.

            Comment


            • #36
              Originally posted by andyprough View Post
              Interesting. systemd OS vs systemd OS, and the winner is ---- no one.

              Comment


              • #37
                Lol they really improved Ubuntu's performance. Few years ago it was the worst distro in case of performance.
                Btw, Arch's kernel is a vanilla kernel with no optimizations, and like other Arch packages, is built with just -O2 for general x86_64.

                Comment


                • #38
                  Originally posted by StarterX4 View Post
                  Lol they really improved Ubuntu's performance. Few years ago it was the worst distro in case of performance.
                  I don't recall Arch ever performing terribly well in the Phoronix multi-distro benchmark contests. This result seems pretty typical.

                  Comment


                  • #39
                    Originally posted by perpetually high View Post
                    You guys are leaving a lot of performance on the table. For what? These changes are all free. And you're on the Phoronix forums. You want the Need for Speed, don't you?

                    You can and will do better. I will see to it.
                    What are you on about? The only thing I can conclude from those microbenchmarks is that the only thing with some effect on performance is Spectre mitigations. Considering the kind of benchmarks you ran I wouldn't expect to see any difference anyway.

                    If you want to actually test task scheduler performance, you need to run a more complex workload. For instance, do a video encoding test with and without some other CPU intensive load running in the background.

                    Comment


                    • #40
                      Originally posted by MadCatX View Post
                      What are you on about? The only thing I can conclude from those microbenchmarks is that the only thing with some effect on performance is Spectre mitigations. Considering the kind of benchmarks you ran I wouldn't expect to see any difference anyway.

                      If you want to actually test task scheduler performance, you need to run a more complex workload. For instance, do a video encoding test with and without some other CPU intensive load running in the background.
                      You’re being a hater. I leave the benchmarks to the pros like Michael. I was just providing some context.

                      And the results are real, bud. My Haswell machine flies. I know when it didn’t. And it does now. A lot has to do with overall GNOME improvements though. Can make any comment you want, I know my machine like the back of my hand.

                      Why in the world would I not take the liberty to compile my own supercharged kernel? Just cause you don’t like it no one else should? Call yourself a Linux pro and don’t even compile your own? But you probably have an 8c/16t. Gtfo.

                      Comment

                      Working...
                      X