Announcement

Collapse
No announcement yet.

Is SteamOS Any Faster Than Ubuntu 15.10 Linux?

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

  • Is SteamOS Any Faster Than Ubuntu 15.10 Linux?

    Phoronix: Is SteamOS Any Faster Than Ubuntu 15.10 Linux?

    Over the past few days have been a number of SteamOS Linux gaming benchmarks, namely published so far are the 22-Way Comparison Of NVIDIA & AMD Graphics Cards On SteamOS For Steam Linux Gaming and 4K AMD/NVIDIA High-End GPU Comparison On SteamOS Linux. When seeing all of those SteamOS results, you may have started wondering: is SteamOS any faster/slower than say Ubuntu Linux? In this article are some benchmarks comparing SteamOS to Ubuntu 15.10.

    Phoronix, Linux Hardware Reviews, Linux hardware benchmarks, Linux server benchmarks, Linux benchmarking, Desktop Linux, Linux performance, Open Source graphics, Linux How To, Ubuntu benchmarks, Ubuntu hardware, Phoronix Test Suite

  • #2
    So, overall, from the four NVIDIA graphics cards tested for this brief comparison,
    SteamOS 2.0 Brewmaster doesn't appear to have much of a performance impact compared to Ubuntu 15.10
    FTFY

    otherwise it's confusing


    How about latency ? min/max framerate between those two ?

    I get best response rate with threadirqs, chrt (raised IRQ priority) tweaks and scheduler tuning or different scheduler (BFS)

    but I doubt that Valve would go that way anytime soon (they are already ?)

    cause sometimes this can cause instability: usb 3.0 and large transfers a few minor kernel releases back - lockup
    Last edited by kernelOfTruth; 27 October 2015, 03:28 PM.

    Comment


    • #3
      Originally posted by kernelOfTruth View Post
      How about latency ? min/max framerate between those two ?
      As explained many times now, the test is limited to what the game engine exposes... With these tests, not much beyond the average FPS. So short of running other PTS non-Steam tests that do expose frame-times or other details, not much can be done.
      Michael Larabel
      https://www.michaellarabel.com/

      Comment


      • #4
        There is a new fglrx update that works around the problems in wily-proposed since today.

        Comment


        • #5
          Originally posted by Michael View Post

          As explained many times now, the test is limited to what the game engine exposes... With these tests, not much beyond the average FPS. So short of running other PTS non-Steam tests that do expose frame-times or other details, not much can be done.

          I know, I know,

          it's just ranting on a high level and also not blaming you,

          perhaps someone comes up with a kind of wrapper that could integrate data from latencytop and combining this with the benchmarks & PTS would provide a more complete picture ...

          Comment


          • #6
            the point of steamos is not the performance, is the easy to use, ppl simply install the game and run it

            Comment


            • #7
              Originally posted by andre30correia View Post
              the point of steamos is not the performance, is the easy to use, ppl simply install the game and run it
              The point of steamos is to give Valve a distribution channel where they don't have to fork over a percentage of their revenue to another company.

              Comment


              • #8
                Originally posted by Michael View Post

                As explained many times now, the test is limited to what the game engine exposes... With these tests, not much beyond the average FPS. So short of running other PTS non-Steam tests that do expose frame-times or other details, not much can be done.
                as it seems benchmarks were run on gnome. you're doing exact same nonsense as benchmarking Gnome on wayland and then proclaiming that as Wayland benchmarks

                if you switch to desktop mode you use different compositor in steamos and get bogus results for anyone interested in running it. in steam it uses its own while in gnome it uses mutter. same as number of running services is very limited in steam session while full blown on desktop. not saying it will sway one way or another, in fact for a long time desktop fared better than steam session, until they fixed the problems.

                if you really want to benchmark steamos, then add benchmark as non steam application and run it from there under steam session.

                Comment


                • #9
                  Good to see the performance identical between SteamOS and Ubuntu. The last thing you want to see is inconsistency in the Linux ecosystem.

                  Comment


                  • #10
                    Is it possible to do a benchmark where the game is pretty much the only thing running as some sort of baseline?

                    Comment

                    Working...
                    X