AMD Ryzen 7 7840U Windows 11 vs. Linux CPU Performance

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts
  • mSparks
    Senior Member
    • Oct 2007
    • 2081

    #41
    Originally posted by avis View Post
    Michael has never mentioned whether he prepares Windows for benchmarking properly.

    And that's a lot of work that needs to be done otherwise you could be getting quite low results.

    https://www.techpowerup.com/forums/t...chmark.287480/
    Does deleting the System32 folder still need doing on windows 11 or did they finally get rid of it?

    Comment

    • sophisticles
      Senior Member
      • Dec 2015
      • 2591

      #42
      Originally posted by andre30correia View Post

      Maybe your laptop is not compatible with Linux, i have no such problems with mine
      It's an HP with an Icelake CPU and Intel graphics. If this isn't compatible, what is?

      Comment

      • sophisticles
        Senior Member
        • Dec 2015
        • 2591

        #43
        Originally posted by xAlt7x View Post
        @sophisticles , @ Sonadow
        Сan we please stay on topic​ and not provoke flame-wars? This article specifically compares Windows and Linux performance on the same hardware.
        Arguments like "Windows might lose some benchmarks but at least it doesn't suck like Linux for X and Y use-cases" seem fanboy-ish to me.
        I'm not trying to provoke a flame war, what I very clearly said was that pure performance is less important than stability.

        Think of a sports car, one car can go 0-60 in 3 sec flat and has a top speed of over 200 mph but breaks down every other month and needs a complete engine overhaul after every track session and the other goes 0-60 in 5 seconds and has a top speed of "only" 160 but the only thing it needs is an oil change and new plugs every 6 months, which is the better car?

        What I think would be interesting is if Michael did an endurance benchmark, like Tom's Hardware did years ago, where they ran 4 very CPU intensive benchmarks simultaneously on a loop for I think it was 48 hours.

        Perhaps Michael could run 4 instances of the PTS benchmark simultaneously a couple of times and see which OS even completes the test much less finishes first.

        I think you may be surprised at the results.

        The point is no one uses an application for 30 seconds at a time, we use our computers for hours at a time and usually we are doing multuple things at the same time, with a number of background tasks running while we work.

        We don't use our computers like a top fuel dragster we use our computers like a Nascar race car, running them nonstop for hours and days and weeks at a time.

        Comment

        • paulocoghi
          Phoronix Member
          • Jul 2019
          • 58

          #44
          Originally posted by avis View Post

          When you do benchmarking you want your system behavior to be predictable and tests to be repeatable.

          If you don't prepare Windows for that, it may run certain background jobs unexpectedly (not limited to Windows Updates) and screw up your testing completely.

          Linux doesn't need that because an average Linux desktop install almost has zero background jobs aside from system logging.
          If the negative impact of those standard Windows's background jobs can "screw up your testing completely", how we could not conclude that these same background jobs will create the same negative impacts and "screw up" the day-today usage "completely".

          Thanks for helping us to understand that, yes, Windows provides negative impacts on its standard form that are so negative, that measures need to be taken to make Windows "benchmarkable".

          Also, thank you to remember us that, yes, the average Linux desktop is so clean that it "install almost zero background jobs" and, thus, provide not only almost zero impact on benchmarks on its standard form, but also almost zero impact on day-to-day usage.

          Thanks for proving, yourself, that Linux is faster than Windows.
          Last edited by paulocoghi; 19 August 2023, 06:42 PM.

          Comment

          • mSparks
            Senior Member
            • Oct 2007
            • 2081

            #45
            Originally posted by sophisticles View Post

            I'm not trying to provoke a flame war, what I very clearly said was that pure performance is less important than stability.

            Think of a sports car, one car can go 0-60 in 3 sec flat and has a top speed of over 200 mph but breaks down every other month and needs a complete engine overhaul after every track session and the other goes 0-60 in 5 seconds and has a top speed of "only" 160 but the only thing it needs is an oil change and new plugs every 6 months, which is the better car?

            What I think would be interesting is if Michael did an endurance benchmark, like Tom's Hardware did years ago, where they ran 4 very CPU intensive benchmarks simultaneously on a loop for I think it was 48 hours.

            Perhaps Michael could run 4 instances of the PTS benchmark simultaneously a couple of times and see which OS even completes the test much less finishes first.

            I think you may be surprised at the results.

            The point is no one uses an application for 30 seconds at a time, we use our computers for hours at a time and usually we are doing multuple things at the same time, with a number of background tasks running while we work.

            We don't use our computers like a top fuel dragster we use our computers like a Nascar race car, running them nonstop for hours and days and weeks at a time.
            Sounds like you dont macOS yet.
            You should. As of 2023 a Linux destop with a high end nvidia gpu and amd cpu with an M2 chip running macOS for a second machine is as good as stable + does everything you could possibly want with insanely high performance can get.

            You also seem to miss the point of this benchmark.
            Linux wins here even if performance was identical - what people want to know is can they buy this laptop chip and expect at least as good as performance as windows it ships with.

            They are probably running a Mac Studio for desktop, which is also a decent setup.

            The fact Linux hands down gets double the performance in some cases is just a bonus/to be expected, this is a 16T CPU, windows generally tops out at 4T for any software that otherwise relies on fork() or atomic Virt Memory. So up to 4x the perf can easily be anticipated (but no benchmarks to show it, cos windows can't fork())

            There is no decent set up that includes windows. Its a steaming pile of clunky under invested shite that is lucky if it can make it through a single days use without exploding.
            Last edited by mSparks; 19 August 2023, 11:14 PM.

            Comment

            • Sonadow
              Senior Member
              • Jun 2009
              • 2278

              #46
              Originally posted by mSparks View Post

              Sounds like you dont macOS yet.
              You should. As of 2023 a Linux destop with a high end nvidia gpu and amd cpu with an M2 chip running macOS for a second machine is as good as stable + does everything you could possibly want with insanely high performance can get.

              You also seem to miss the point of this benchmark.
              Linux wins here even if performance was identical - what people want to know is can they buy this laptop chip and expect at least as good as performance as windows it ships with.

              They are probably running a Mac Studio for desktop, which is also a decent setup.

              The fact Linux hands down gets double the performance in some cases is just a bonus/to be expected, this is a 16T CPU, windows generally tops out at 4T for any software that otherwise relies on fork() or atomic Virt Memory. So up to 4x the perf can easily be anticipated (but no benchmarks to show it, cos windows can't fork())

              There is no decent set up that includes windows. Its a steaming pile of clunky under invested shite that is lucky if it can make it through a single days use without exploding.
              Sure, keep yapping all you want when many software, applications and drivers used by billions of people all over the world and in production systems only exist on Windows and no where else, not even Linux.

              The fastest Linux distribution means jack shit when it can't even talk to a specific piece of hardware. Or run a specific workload.

              And that 4T is just complete bullshit.
              Last edited by Sonadow; 20 August 2023, 09:21 AM.

              Comment

              • avis
                Senior Member
                • Dec 2022
                • 2252

                #47
                Originally posted by xAlt7x View Post
                By that logic we should tweak Linux as well (turning off services like Gnome Tracker, Baloo). And although this sounds good in theory, but in practice​ it will to lead to inconsistent results when someone tries to repeat Michael's benchmarks.
                This makes perfect sense however these services are built on top of the inotify/fnotify subsystem, so they normally should be completely idle.

                Comment

                • avis
                  Senior Member
                  • Dec 2022
                  • 2252

                  #48
                  Originally posted by paulocoghi View Post

                  If the negative impact of those standard Windows's background jobs can "screw up your testing completely", how we could not conclude that these same background jobs will create the same negative impacts and "screw up" the day-today usage "completely".

                  Thanks for helping us to understand that, yes, Windows provides negative impacts on its standard form that are so negative, that measures need to be taken to make Windows "benchmarkable".

                  Also, thank you to remember us that, yes, the average Linux desktop is so clean that it "install almost zero background jobs" and, thus, provide not only almost zero impact on benchmarks on its standard form, but also almost zero impact on day-to-day usage.

                  Thanks for proving, yourself, that Linux is faster than Windows.
                  If ~10% or so higher performance is everything you need from the OS I've got bad news for you: the Linux market share on desktop remains in low single digits.

                  People choose something which works better, not something that performs marginally better. But yeah, you can boast about Linux performance to your Windows friends. You know what they'll do? They will simply shrug off your achievements.

                  Comment

                  • Sonadow
                    Senior Member
                    • Jun 2009
                    • 2278

                    #49
                    Originally posted by sophisticles View Post

                    It's an HP with an Icelake CPU and Intel graphics. If this isn't compatible, what is?
                    IceLake works *very* well on a clean ISO installation of Windows 11 downloaded from Microsoft's website. Speaking from experience.

                    However, in all my installations I artificially gimp the performance by forcing Windows to lock the CPU's processor to the lowest possible speed since all notebooks have shit-tier cooling.

                    As far as Linux is concerned, I won't run Icelake on any kernel below 5.16.
                    Last edited by Sonadow; 20 August 2023, 09:12 AM.

                    Comment

                    • HEL88
                      Senior Member
                      • Oct 2020
                      • 412

                      #50
                      Originally posted by schmidtbag View Post
                      It's more about day to day use and responsiveness. When you're just crunching numbers, the OS hardly makes a difference; the scheduler matters more. Where Linux tends to greatly outpace Windows is things like installations/updates, cold booting, and loading basic functions (like the Start menu or Task Manager). Windows carries a lot of bloat, which means it takes longer for it to swap things in/out of memory, which also has to happen more often since memory fills up quicker.
                      So, why bloating Windows with poore scheduler is faster than linux in:

                      phoronix.com/benchmark/result/amd-ryzen-7-7840u-benchmark-windows-vs-linux/luxcorerender-luxcore-benchmark-cpu.svgz

                      phoronix.com/benchmark/result/amd-ryzen-7-7840u-benchmark-windows-vs-linux/kvazaar-bosphorus-4k-medium.svgz
                      phoronix.com/benchmark/result/amd-ryzen-7-7840u-benchmark-windows-vs-linux/kvazaar-bosphorus-1080p-medium.svgz
                      phoronix.com/benchmark/result/amd-ryzen-7-7840u-benchmark-windows-vs-linux/kvazaar-bosphorus-4k-very-fast.svgz
                      phoronix.com/benchmark/result/amd-ryzen-7-7840u-benchmark-windows-vs-linux/kvazaar-bosphorus-1080p-very-fast.svgz

                      phoronix.com/benchmark/result/amd-ryzen-7-7840u-benchmark-windows-vs-linux/intel-open-image-denoise-rthdr_alb_nrm3840x2160-cpu-only.svgz
                      phoronix.com/benchmark/result/amd-ryzen-7-7840u-benchmark-windows-vs-linux/intel-open-image-denoise-rthdr_alb_nrm3840x2160-cpu-only.svgz
                      phoronix.com/benchmark/result/amd-ryzen-7-7840u-benchmark-windows-vs-linux/intel-open-image-denoise-rtldr_alb_nrm3840x2160-cpu-only.svgz
                      phoronix.com/benchmark/result/amd-ryzen-7-7840u-benchmark-windows-vs-linux/intel-open-image-denoise-rtlightmaphdr4096x4096-cpu-only.svgz

                      phoronix.com/benchmark/result/amd-ryzen-7-7840u-benchmark-windows-vs-linux/ospray-particle_volume-scivis-real_time.svgz
                      phoronix.com/benchmark/result/amd-ryzen-7-7840u-benchmark-windows-vs-linux/ospray-particle_volume-ao-real_time.svgz

                      phoronix.com/benchmark/result/amd-ryzen-7-7840u-benchmark-windows-vs-linux/primesieve-1e12.svgz

                      phoronix.com/benchmark/result/amd-ryzen-7-7840u-benchmark-windows-vs-linux/y-cruncher-500m.svgz

                      phoronix.com/benchmark/result/amd-ryzen-7-7840u-benchmark-windows-vs-linux/astc-encoder-thorough.svgz

                      cold booting,
                      My windows booting very fast, similar to linux. But I reboot my Windows once every few weeks sometimes once a week. I use sleep - like most people.

                      Linux I understand is less stable so it needs to be restarted often and cold boot time is important?​

                      Windows is things like installations/updates,
                      An update that requires a reboot comes once every few weeks.

                      On windows, I turn on the update - something is doing in the background. I reboot when I go to get food or go to the toilet. When I come back, it's ready to go to work.

                      I understand that because you have to restart your computer once every few weeks, this system is not suitable for work?

                      installations
                      My desktop Windows is about 5-6 years old, it survived the change of disk and motherboard and is doing great.

                      On the laptop Windows 11 is also from new and I do not rewind some 'installation'.

                      Windows is not Linux that you have to reinstall every 3 months.

                      Comment

                      Working...
                      X