Announcement

Collapse
No announcement yet.

Preview: OS X 10.10 Yosemite vs. Ubuntu Linux GPU Performance

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

  • Preview: OS X 10.10 Yosemite vs. Ubuntu Linux GPU Performance

    Phoronix: Preview: OS X 10.10 Yosemite vs. Ubuntu Linux GPU Performance

    At the request of many Phoronix readers, here's our first tests of Apple's OS X 10.10 "Yosemite" operating system as we see how the OpenGL performance compares between it and Ubuntu Linux with an updated kernel and Mesa.

    http://www.phoronix.com/vr.php?view=20802

  • #2
    run some computacional benchmarks like h264 encoding etc and some browser javascript and html5 speed benchmarks


    testing opengl games you arejust measuring intel's drivers and little else

    Comment


    • #3
      Originally posted by Pallidus View Post
      run some computacional benchmarks like h264 encoding etc and some browser javascript and html5 speed benchmarks


      testing opengl games you arejust measuring intel's drivers and little else
      That's a good point but I'm not sure I want to see the results for that. I have a feeling Mac is going to destroy linux in that perspective; it might even outperform Windows.

      Comment


      • #4
        not according to this guy and other sources I've seen:

        http://grigio.org/mac_osx_10_10_yose...cbook_air_2011


        I would really like phoronix to validate this results but they seem to test opengl and nothing else, time and time again....


        Listen opengl performance is a direct result of intel not giving a shit about their linux driver's and little else can be extrapolated from that... run computational benchmarks like you have done while comparing bsd's to linux...

        Comment


        • #5
          mac osx

          Originally posted by schmidtbag View Post
          That's a good point but I'm not sure I want to see the results for that. I have a feeling Mac is going to destroy linux in that perspective; it might even outperform Windows.
          mac osx is the worst in comput benchs is well known

          Comment


          • #6
            Originally posted by rikkinho View Post
            mac osx is the worst in comput benchs is well known
            I think it depends what kind. For things like video playback and we browser stuff, I'm sure Mac is pretty good.

            Comment


            • #7
              Originally posted by schmidtbag View Post
              I think it depends what kind. For things like video playback and we browser stuff, I'm sure Mac is pretty good.
              It is not. It is slow as mollases. It always was, it will be for ever. Check older benchmarks in the specific site and you'll see. OSX is even slower than windows.
              OSX used to be faster than Linux in games. But now since GPU drivers in Linux are in much better shape than 2-3 years before, it is faster at nothing.

              Comment


              • #8
                Originally posted by Apopas View Post
                It is not. It is slow as mollases. It always was, it will be for ever. Check older benchmarks in the specific site and you'll see. OSX is even slower than windows.
                OSX used to be faster than Linux in games. But now since GPU drivers in Linux are in much better shape than 2-3 years before, it is faster at nothing.
                I don't know, it seems to be pretty fast for me. I would bet that my OS X machine would outperform your Linux / Windows machine in every benchmark. Mind you, not running server applications, but computational benchmarks. I'm not actually going to run PTS because I refuse to have PHP installed in any way as PHP is cancer, but I have a hex-core i7, 16GB of RAM, and a pretty solid SSD with a GTX 680. VMware Fusion outperforms KVM in my practical uses and features (VMware Workstation does exist thankfully), and I'm sure that if I had Linux on my hardware I would get slightly faster disk reads and slightly better CPU scheduling, OS X still has the features I care about and it JUST WORKS. I don't have to configure it, I just turn it on, do work, then turn it off. I have yet to have one instance where it even considered getting slow.
                Last edited by jimbohale; 08-24-2014, 05:48 PM.

                Comment


                • #9
                  Originally posted by schmidtbag View Post
                  That's a good point but I'm not sure I want to see the results for that. I have a feeling Mac is going to destroy linux in that perspective; it might even outperform Windows.
                  I wouldn't expect that at all (at least if you use chrome/chromium---ff is still using cairo on linux though you can enable skia on nightlies).
                  And the computational stuff typically does best on linux.
                  Personally, I'd love to see some native 2d benchmarks.

                  Comment


                  • #10
                    Doubt that

                    Originally posted by jimbohale View Post
                    I don't know, it seems to be pretty fast for me. I would bet that my OS X machine would outperform your Linux / Windows machine in every benchmark. Mind you, not running server applications, but computational benchmarks. I'm not actually going to run PTS because I refuse to have PHP installed in any way as PHP is cancer, but I have a hex-core i7, 16GB of RAM, and a pretty solid SSD with a GTX 680. VMware Fusion outperforms KVM in my practical uses and features (VMware Workstation does exist thankfully), and I'm sure that if I had Linux on my hardware I would get slightly faster disk reads and slightly better CPU scheduling, OS X still has the features I care about and it JUST WORKS. I don't have to configure it, I just turn it on, do work, then turn it off. I have yet to have one instance where it even considered getting slow.

                    I really doubt that. I agree that AQUA widgets and the whole environment itself is very responsive, and animaitons are smooth as butter, but in the end it comes down to the applications you are going to use. A chrome is not going to be faster under mac than under linux, or firefox, or other applications.

                    And for linux in this category we got elementary OS, which is nearly as fast as AQUA widgets and environment.

                    I doubt that Mac will win in benchmarks, as Linux gets new graphics update all the time, and Mac OS X not. And also you got much more flexibility. There were some older benchmarks regarding intel on linux and mac, and linux won at nearly every point ;X

                    But still, Mac OS X is the best UNIX I know off... (also taking linux into this category)

                    Comment


                    • #11
                      Originally posted by gotwig View Post
                      I really doubt that. I agree that AQUA widgets and the whole environment itself is very responsive, and animaitons are smooth as butter, but in the end it comes down to the applications you are going to use. A chrome is not going to be faster under mac than under linux, or firefox, or other applications.

                      And for linux in this category we got elementary OS, which is nearly as fast as AQUA widgets and environment.

                      I doubt that Mac will win in benchmarks, as Linux gets new graphics update all the time, and Mac OS X not. And also you got much more flexibility. There were some older benchmarks regarding intel on linux and mac, and linux won at nearly every point ;X

                      But still, Mac OS X is the best UNIX I know off... (also taking linux into this category)
                      http://www.tomshardware.com/reviews/...u,3129-18.html

                      Also, linux isn't a unix

                      Comment


                      • #12
                        Originally posted by jimbohale View Post
                        I don't know, it seems to be pretty fast for me. I would bet that my OS X machine would outperform your Linux / Windows machine in every benchmark. Mind you, not running server applications, but computational benchmarks. I'm not actually going to run PTS because I refuse to have PHP installed in any way as PHP is cancer, but I have a hex-core i7, 16GB of RAM, and a pretty solid SSD with a GTX 680. VMware Fusion outperforms KVM in my practical uses and features (VMware Workstation does exist thankfully), and I'm sure that if I had Linux on my hardware I would get slightly faster disk reads and slightly better CPU scheduling,
                        To be sure without testing is a bit dumb, don't you think? I'm telling you because I've tried both.
                        And yes in your machine even windows Vista would feel fast but our point is not that but how it would perform against another OS.
                        Check here http://www.phoronix.com/scan.php?pag...4_ubuntu&num=4
                        And keep in mind that it is against Ubuntu and in a OSX optimized machine. Hell, if you throw Arch or my Gentoo on the table, the results are gonna be even more catastrophic for OSX. And as I said before, check and even older benchmarks, the same story.
                        OS X still has the features I care about and it JUST WORKS. I don't have to configure it, I just turn it on, do work, then turn it off. I have yet to have one instance where it even considered getting slow.
                        Well we are talking about performance here, not why you prefer OSX. "Just works" means absolutely nothing for me. Seriously. Gentoo user here.

                        Comment


                        • #13


                          This looks like to me game is capped at 30 fps

                          Comment


                          • #14
                            Here is the big question in my mind: there are a few opengl games where Mac OS X consistently out performs linux and there are some games that the reverse is true. What is sub-optimal with the Linux driver? Is it the same code path that is bottle necking the performance? Is it different code paths?

                            Comment


                            • #15
                              Most biased benchmark:
                              -No enough hardware diversity.
                              - No enough games, mostly not games using the GPU juice.

                              Please provide a proper review, this is beyond ridicule and manipulation. Michael Larravel, you can do it a lot better with the hardware you own.

                              Comment

                              Working...
                              X