Announcement

Collapse
No announcement yet.

Is Windows 7 Actually Faster Than Ubuntu 10.04?

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

  • #46
    Originally posted by hax0r View Post
    Windows is still a lot better for desktop, where latency and GUI response matters. There's is no good DE today, none of them are optimized or are memory efficient. They're are slow, featureless, and they suck.
    Windows might be better for a desktop than Ubuntu Linux, but it is by no means better than Linux. Ubuntu Linux has kernel that is optimized for servers, which prevents a fair comparison from being done of Linux and Windows. Just the fact that Linux is represented by Ubuntu does not mean that it is a fair representation of what Linux can actually do.

    Comment


    • #47
      I'm setting up the PTS on Win7 right now myself, and I see that UT2004 isn't integrated into it yet, so I can understand why it wasn't included in the test then. When will it be implemented, by the way?
      Also, a test based on Unreal Engine 3 would be awesome since a lot of current day games run on it... However, I'm not aware of any UE3 games on Linux, although it's obviously possible (we even saw UT3 screenshots on Linux...).

      Comment


      • #48
        Originally posted by kraftman View Post

        Scheduler performance (average, worse):

        Linux: 0.009mS 0.3mS
        Windows: 2mS scheduling latency 16mS

        http://widefox.pbworks.com/Scheduler#Timeslice

        Windows looks like a big, fat cow. If you cannot backup what you said just be a nice troll and be quiet, please.
        With all due respect Kraftman that link is a bit dated. We know for a fact that there has been performance regressions introduced since 2.6.28 and that Win 7 has improved it's performance over Vista so that link isn't exactly representative of current offerings.

        Comment


        • #49
          Originally posted by deanjo View Post
          One other thing I would suggest, especially on the gaming tests is to run them with sound enabled as this could have measurable effects on game play.
          Yeah, especially under Gnome with Pulse Audio enabled.

          Comment


          • #50
            Originally posted by deanjo View Post
            With all due respect Kraftman that link is a bit dated. We know for a fact that there has been performance regressions introduced since 2.6.28 and that Win 7 has improved it's performance over Vista so that link isn't exactly representative of current offerings.
            Which and where (in both cases Linux and Windows)?

            Comment


            • #51
              Originally posted by Shining Arcanine View Post
              Ubuntu Linux has kernel that is optimized for servers, which prevents a fair comparison from being done of Linux and Windows.
              Not really sure where you get the idea that Ubuntu's default kernel has been optimized for servers. The kernel configs in Ubuntu are aimed as a multi-role config. It's not optimized specifically for server or desktop use. If you want a "in the can" default desktop kernel take a look at openSUSE where it defaults to a kernel that is "desktop tuned".

              Comment


              • #52
                @Deanjo

                There were slowdowns in Ext4 due to some changes, but this is about scheduler latency. Btw. Phoronix become so unresponsive it's hard to edit a post in time.

                Comment


                • #53
                  @Deanjo

                  There were slowdowns in Ext4 due to some changes, but this is about scheduler latency. After BFS appeared the CFS became even better. Btw. Phoronix is so unresponsive now it's hard to edit a post in time.

                  Comment


                  • #54
                    The fact is that Linux is at the end of the feeding chain, so don't expect any big FPS in games, that's Windows world. This days will found some morons that will install Ubuntu on 700$+ system that's stupid. Linux can be very good for workstation and cheap one, that's it.

                    Comment


                    • #55
                      Originally posted by kraftman View Post
                      @Deanjo

                      There were slowdowns in Ext4 due to some changes, but this is about scheduler latency. Btw. Phoronix become so unresponsive it's hard to edit a post in time.
                      Ya, don't know what is up with the server. But going back the scheduler improvements found in Win 7

                      http://blogs.msdn.com/philpenn/archi...rallelism.aspx

                      as well as other improvements

                      http://www.microsoft.com/whdc/system...TimerCoal.mspx

                      http://msdn.microsoft.com/en-us/libr...8VS.85%29.aspx

                      These were not present in Vista.

                      Of course there are many other factors which effect the end users experience as to what is responsive to them and what is not, such as pulseaudio (or really audio as a whole in linux), power management, graphics, etc etc. Yes I'm aware that most of the regressions come from the Ext4 filesystem but that has to be taken into account when talking about the end user experience.

                      Comment


                      • #56
                        Originally posted by deanjo View Post
                        Ya, don't know what is up with the server. But going back the scheduler improvements found in Win 7

                        http://blogs.msdn.com/philpenn/archi...rallelism.aspx

                        as well as other improvements

                        http://www.microsoft.com/whdc/system...TimerCoal.mspx

                        http://msdn.microsoft.com/en-us/libr...8VS.85%29.aspx

                        These were not present in Vista.
                        Good to see they're improving, but we don't know how well this improved latencies. While Linux kernel is probably more responsive (there's also RT Linux), more important for desktops are visible latencies which are sometimes caused by slow 2D.

                        Of course there are many other factors which effect the end users experience as to what is responsive to them and what is not, such as pulseaudio (or really audio as a whole in linux), power management, graphics, etc etc.
                        Yes, sometimes there are problems with this, but it depends on configurations.

                        Yes I'm aware that most of the regressions come from the Ext4 filesystem but that has to be taken into account when talking about the end user experience.
                        It depends if end users suffer from Ext4 changes, but they/we probably do.

                        @Filip007

                        The fact is that Linux is at the end of the feeding chain, so don't expect any big FPS in games, that's Windows world. This days will found some morons that will install Ubuntu on 700$+ system that's stupid. Linux can be very good for workstation and cheap one, that's it.
                        Linux is also very good for desktops. What's wrong with those FPS numbers (sometimes better then in Windows)? Games are very important for desktops, but there are also other things. Some people even prefer playing under Wine then keeping Windows on a hard drive.

                        Comment


                        • #57
                          Originally posted by kraftman View Post
                          Linux is also very good for desktops. What's wrong with those FPS numbers (sometimes better then in Windows)? Games are very important for desktops, but there are also other things. Some people even prefer playing under Wine then keeping Windows on a hard drive.
                          Indeed. If Wine solved their problem with DirectInput that makes more than 360 degree turns in UE-based games impossible, I'd make Linux my first OS. Unfortunately, they are still having trouble with latest XInput versions...

                          Comment


                          • #58
                            For some reason, some people seem to interpret the results as if Windows has won this benchmark big time. Is this because the title and the intro suggest that Windows might win, and the readers don't bother to read further?

                            Comment


                            • #59
                              Originally posted by Remco View Post
                              For some reason, some people seem to interpret the results as if Windows has won this benchmark big time. Is this because the title and the intro suggest that Windows might win, and the readers don't bother to read further?
                              This is for sure the case

                              Comment


                              • #60
                                Nice work, Michael!

                                The most important results are without doubt those of XPlane, Unigine and SPECviewperf, and Linux makes a strong showing in all. Specific anomalies aside, performance results are very similar to what I am observing in my own (non-PTS) tests.

                                We can safely ignore any test that exceeds 200fps or so. Performance differences above that point are meaningless and reveal little about real-world performance in modern software (they are equivalent to glxgears in that regard).

                                Finally, please don't feed the Ubuntu-is-not-Linux trolls.

                                Comment

                                Working...
                                X