Announcement

Collapse
No announcement yet.

NVIDIA GTX 680: Windows 7 vs. Ubuntu 12.04

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

  • #11
    Originally posted by FourDMusic View Post
    While the performance seems to be comparable, I wonder how much performance improvement we'll see using Wayland. Will this push the linux graphics stack past windows? It'll be nice having a display server without all the X11 cruft that was based on 1980's graphics cards either way.

    Too bad Nvidia doesn't play to support the protocol... yet.
    GLX direct rendering does not go through the X11 server.

    Comment


    • #12
      More than gaming performance, I would like to know what's with cuda/opencl performance vs other nvidia & ati gfx cards.

      Comment


      • #13
        Originally posted by Milyardo View Post
        X.Org isn't the reason 3D performance is poor on open source drivers anyways. Mesa is the problem with 3D graphics stack. X.Org does have some issues, like in events and input, things that make compositing difficult. Things like running multiple drivers at the same time, and GPGPU computing are difficult because of X. X.Org nor X11 impose any significant limitation on the performance 3D games.
        That's interesting and I'm sure you're right about that but you're missing 1 key point - xorg is relatively a mess of code at this point, which is (AFAIK) the reason why wayland was created in the first place. The extra clutter in X makes it more difficult to code for compared to wayland, at least that's the impression I've been getting. For something as complex as GPUs (which IMO are arguably the most complex hardware), anything that can make development more difficult and hindered will probably result in a less-than-perfect result. Since Windows probably gets more attention, doesn't have variations of DEs, and is likely a lot more consistent than X, its no surprise why it generally performs better.

        Comment


        • #14
          Originally posted by Veerappan View Post
          Care to explain? I've successfully used the Catalyst drivers + APPSDK for OpenCL development on both Windows 7 and Ubuntu (11.04 and 11.10 at least). Did they break something in the newest versions that I haven't heard about?
          The problem with OpenCL implementation in Cycles is described here:

          Comment


          • #15
            Yea, it makes sense to compare stock Win7 to stock Ubuntu. Although it would be very interesting to see how it compares to optimised Win7 and optimised Ubuntu (TWM and such ).

            Comment


            • #16
              Originally posted by gilboa View Post
              ... And this with Unity 3D.

              Maybe in 2019 Phoronix will ceases to compare the slow(est) [1] WM [2] to Windows and OSX.
              *Sigh*

              - Gilboa
              [1]http://www.phoronix.com/scan.php?page=article&item=linux_desktop_managers1 &num=4
              [2] http://www.phoronix.com/scan.php?pag...desktops&num=2

              Originally posted by GreatEmerald View Post
              Yea, it makes sense to compare stock Win7 to stock Ubuntu. Although it would be very interesting to see how it compares to optimised Win7 and optimised Ubuntu (TWM and such ).
              +1
              Who cares of the window manager when the game is fullscreen?
              PTS should support benchmarking in a separate X server, totally blank and without even a window manager.
              I'm looking forward for year 2019...

              Comment


              • #17
                Originally posted by Teho View Post
                Ubuntu is the most popular desktop Linux distribution so comparing to it is only natural and it just happen to use Unity.
                Given the fact that no-one ever bothers to methodically sample Linux users (aside from Distrowatch) and given the inherit difficulty in trying to measure the user-base of freely distributed operating system - I *fail* to see how anyone can magically decided that Ubuntu *is* Linux and as a result Unity *is* the most widely used DE.

                Speaking for distrowatch (again the only reliable / non-reliable source of Linux userbase statistics) in the past six months, *not* only Ubuntu was second base, but it also accounted for *only* ~20% of the total user base. *

                .... And even if we agree that Ubuntu is that yardstick, which not also add Kubuntu? Xubutnu?
                In my small little world, I'd first find the fastest Linux distribution and *than* compare it to OSX and Windows 7... But guess its just me.

                - Gilboa
                * Granted Mint is based on Ubuntu, but A. AFAIR it doesn't use Unity by default and B. Ubunutu is partially based on Debian so I'm not sure it's relevant.
                oVirt-HV1: Intel S2600C0, 2xE5-2658V2, 128GB, 8x2TB, 4x480GB SSD, GTX1080 (to-VM), Dell U3219Q, U2415, U2412M.
                oVirt-HV2: Intel S2400GP2, 2xE5-2448L, 120GB, 8x2TB, 4x480GB SSD, GTX730 (to-VM).
                oVirt-HV3: Gigabyte B85M-HD3, E3-1245V3, 32GB, 4x1TB, 2x480GB SSD, GTX980 (to-VM).
                Devel-2: Asus H110M-K, i5-6500, 16GB, 3x1TB + 128GB-SSD, F33.

                Comment


                • #18
                  Ubuntu themselves claim 20M users (was that the right number?), while other distros don't give numbers generally. Even though they don't give out the method, it's pretty much the only number available.

                  Comment


                  • #19
                    Originally posted by curaga View Post
                    Ubuntu themselves claim 20M users (was that the right number?), while other distros don't give numbers generally. Even though they don't give out the method, it's pretty much the only number available.
                    Somehow I don't put much trust in distro supplied numbers, unless its actual paying customers (E.g. OpenSUSE / RedHat)

                    - Gilboa
                    oVirt-HV1: Intel S2600C0, 2xE5-2658V2, 128GB, 8x2TB, 4x480GB SSD, GTX1080 (to-VM), Dell U3219Q, U2415, U2412M.
                    oVirt-HV2: Intel S2400GP2, 2xE5-2448L, 120GB, 8x2TB, 4x480GB SSD, GTX730 (to-VM).
                    oVirt-HV3: Gigabyte B85M-HD3, E3-1245V3, 32GB, 4x1TB, 2x480GB SSD, GTX980 (to-VM).
                    Devel-2: Asus H110M-K, i5-6500, 16GB, 3x1TB + 128GB-SSD, F33.

                    Comment


                    • #20
                      Originally posted by gilboa View Post
                      Given the fact that no-one ever bothers to methodically sample Linux users (aside from Distrowatch) and given the inherit difficulty in trying to measure the user-base of freely distributed operating system - I *fail* to see how anyone can magically decided that Ubuntu *is* Linux and as a result Unity *is* the most widely used DE.
                      No one here's deciding what is Linux and what is not. This was a comparson between Windows 7 and Ubuntu 12.04 not Windows 7 and Linux.


                      Originally posted by gilboa View Post
                      Speaking for distrowatch (again the only reliable / non-reliable source of Linux userbase statistics) in the past six months, *not* only Ubuntu was second base, but it also accounted for *only* ~20% of the total user base. *
                      There's nothing realible in Distrowatch. None. If you want something that gives even some hint of what popular and what is not then Google Trends is a good start:


                      Originally posted by gilboa View Post
                      .... And even if we agree that Ubuntu is that yardstick, which not also add Kubuntu? Xubutnu?
                      I guess it is fun to test the difference between various Ubuntu spins but this test is not about that. You do understand that it takes time to do tests and those distributions have only a fraction of Ubuntu's userbase.

                      Originally posted by gilboa View Post
                      In my small little world, I'd first find the fastest Linux distribution and *than* compare it to OSX and Windows 7... But guess its just me.
                      What exactly would those results tell us? I mean I think it's a bit more valuable information to know how actually useful platforms perfom than something that is tweaked for just one setup or used by no one.

                      Comment

                      Working...
                      X