No announcement yet.

A Look At The Windows 10 vs. Linux Performance On AMD Threadripper 2990WX

  • Filter
  • Time
  • Show
Clear All
new posts

  • A Look At The Windows 10 vs. Linux Performance On AMD Threadripper 2990WX

    Phoronix: A Look At The Windows 10 vs. Linux Performance On AMD Threadripper 2990WX

    Complementing the extensive Linux benchmarks done earlier today of the AMD Threadripper 2990WX in our review (as well as on the Threadripper 2950X), in this article are our first Windows 10 vs. Linux benchmarks of this 32-core / 64-thread $1799 USD processor. Tests were done from Microsoft Windows 10 against Clear Linux, Ubuntu 18.04, the Arch-based Antergos 18.7-Rolling, and openSUSE Tumbleweed.

  • #2
    What a hilarious indictment of Windows!

    They've been catching up in a number of areas, and I've started to find it a bit more difficult to persuade people to go Linux, but now we see that it's all sugar coating of the bitter pill that the foundations are still decrepit.

    There's no replacement for displacement...

    (and I speak as a relative newcomer to Linux - past 5 years).
    Last edited by vegabook; 13 August 2018, 04:25 PM.


    • #3
      Since MS had to loose some license constrains on Windows for Threadripper, I imagine they forgot to loose other stuff too. Lets see if those performance problems remain on Windows server.

      Anyway, always nice to see Linux wipe the floor with the patent troll.


      • #4
        What makes x264 not scale well on heavy multi-core processors?

        Originally posted by phoronix View Post
        but the Microsoft OS did come out ahead in the Solitaire sub-test.
        I see what you did there...


        • #5
          Can you please also test Blender's Gooseberry scene? It is used in many Windows oriented 2990WX reviews and shows bad scaling over 2950X (only 20% better). I am sure on Linux it will be much much better.


          • #6
            It might be worth pointing out that openSuSE Tumbleweed still uses the (older) SLAB memory allocator, while most other distros stick with the linux default choice of SLUB. It would be interesting to investigate how SLAB performs vs SLUB on these large NUMA systems.


            • #7
              Very interesting results. I'm quite interested by that 7zip compression benchmark. It's good to know that it's not a problem with the CPU itself since 7zip manages to turn in good results under Linux.


              • #8
                this benchmark results let me wondering about all those window benchmarks being released today by several "tech" youtubers. Most of them are comparing new TR results to Intel i9-7980XE on Windows. And even though the results still favor the 32-core TR, most of them dont show that much excitement for the results as expected. It would be nice if more of them tested the platforms on Linux (I know that's asking too much, but still), or at least knew of this big difference and mention it in the review.
                Last edited by euler271; 13 August 2018, 05:49 PM.


                • #9
                  Not only Youtubers test on Windows, virtually the whole tech reviewer community tested on Windows exclusively.
                  And many concluded that Threadripper problems were due to memory problems, and only few blamed Windows for it.

                  I mean, look at this:

                  and compare to what Phoronix found:


                  • #10
                    haasn this probably confirm what you are saying