Announcement

Collapse
No announcement yet.

Arch Linux, Clear Linux & Ubuntu Against Windows 10/11 On Intel Rocket Lake

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

  • Arch Linux, Clear Linux & Ubuntu Against Windows 10/11 On Intel Rocket Lake

    Phoronix: Arch Linux, Clear Linux & Ubuntu Against Windows 10/11 On Intel Rocket Lake

    Last month after Microsoft began publishing their Windows Insider Preview builds of Windows 11, I ran some early Windows 11 benchmarks against Linux using an AMD Ryzen 9 5950X. Linux led in those benchmarks on the AMD Zen 3 desktop while for those wondering if that still holds true for Intel hardware, here are benchmarks of a Core i9 11900K "Rocket Lake" desktop with Windows 10 21H1, Windows 11 in its latest preview build as of testing, and then compared to Arch Linux / Clear Linux / Ubuntu.

    https://www.phoronix.com/vr.php?view=30426

  • #2
    From personal experience I'm pretty sure that the differences observed with Blender during CPU rendering are almost entirely due to the different compilers and toolchain used for building the application. You can try the same rendering test in a virtual machine or also WSL under Windows and obtain almost the same performance as the native Linux build.

    Other cases where large differences are observed may also be due to similar reasons.

    Comment


    • #3
      Originally posted by Solid State Brain View Post
      From personal experience I'm pretty sure that the differences observed with Blender during CPU rendering are almost entirely due to the different compilers and toolchain used for building the application. You can try the same rendering test in a virtual machine or also WSL under Windows and obtain almost the same performance as the native Linux build.

      Other cases where large differences are observed may also be due to similar reasons.
      That, along with the scheduler, power profile, and background tasks soaking up resources.

      Comment


      • #4
        Originally posted by schmidtbag View Post
        That, along with the scheduler, power profile, and background tasks soaking up resources.
        In a virtual machine with Windows as the host there should be in total even more background tasks soaking up resources, yet Blender CPU rendering performance is still higher by about the same percentage observed in the tests here on Phoronix.

        Comment


        • #5
          Can anyone tell me how to view those benchmarks?

          Comment


          • #6
            How about firing up a VM with A: Linux and B: Windows. Then on the host measure CPU usage when A is idle for a couple of minutes. Then the same for B. Now that a baseline is established run a benchmark on A then on B and look at the hosts CPU usage. That will show which OS is the most taxing for the host.... Then do the same with another os for the host and see what happens.

            http://www.dirtcellar.net

            Comment


            • #7
              This benchmark contest is good for looking at some technological differences.

              However, this lineup of distros is not reflective of Windows' actual competition from desktop distros. Literally zero people are going to be abandoning Windows 10 or 11 for Clear Linux, which no longer even has an officially supported desktop product.

              Ultimately, testing Windows 10 & 11 against popular distros for ex-Windows users would be interesting. For example, Mint, Manjaro, and MX are more designed for those leaving a Windows environment, and tend to be more popular than their parent distros (Ubuntu, Arch and Debian) among people leaving the Windows ecosystem. They also tend to have configurations that are more optimized out of the box for gaming and casual computer use than most other distros.

              Comment


              • #8
                Originally posted by andyprough View Post
                This benchmark contest is good for looking at some technological differences.

                However, this lineup of distros is not reflective of Windows' actual competition from desktop distros. Literally zero people are going to be abandoning Windows 10 or 11 for Clear Linux, which no longer even has an officially supported desktop product.

                Ultimately, testing Windows 10 & 11 against popular distros for ex-Windows users would be interesting. For example, Mint, Manjaro, and MX are more designed for those leaving a Windows environment, and tend to be more popular than their parent distros (Ubuntu, Arch and Debian) among people leaving the Windows ecosystem. They also tend to have configurations that are more optimized out of the box for gaming and casual computer use than most other distros.
                For sure, benchmarks are more meaningful when run on an OS that people actually use. Clear Linux is more of a tech demo than a consumer product.

                Comment


                • #9
                  Originally posted by andyprough View Post
                  This benchmark contest is good for looking at some technological differences.

                  However, this lineup of distros is not reflective of Windows' actual competition from desktop distros. Literally zero people are going to be abandoning Windows 10 or 11 for Clear Linux, which no longer even has an officially supported desktop product.

                  Ultimately, testing Windows 10 & 11 against popular distros for ex-Windows users would be interesting. For example, Mint, Manjaro, and MX are more designed for those leaving a Windows environment, and tend to be more popular than their parent distros (Ubuntu, Arch and Debian) among people leaving the Windows ecosystem. They also tend to have configurations that are more optimized out of the box for gaming and casual computer use than most other distros.
                  If SteamOS 3.0 can also comfortably serve as a general-purpose desktop system, then I actually expect it to become the go-to choice for most Windows users considering a switch over to Linux.

                  Hopefully Valve can tackle Arch's greatest weakness (stability after updates) with proper & thorough testing of their OS upgrades, which I expect to happen about once every month.

                  And even though most Windows folks won't make the jump, the sheer existence of the Steam Deck should be enough to make a sizable dent in the statistics over time, and especially so after other third-party manufacturers decide to begin producing similar devices shipping with SteamOS by default, because unlike Microsoft, Valve won't ask for any royalties for their OS, which should make this decision a no-brainer for every OEM out there.

                  Comment


                  • #10
                    Originally posted by Oshank Kashyap View Post
                    Can anyone tell me how to view those benchmarks?
                    Are you serious?

                    There is a button that reads Next Page at the bottom of the article (unless you are a premium user). If you click on it you will see the benchmarks.
                    Now if you mean the OpenBenchmarking.org result, click here.

                    Comment

                    Working...
                    X