Announcement

Collapse
No announcement yet.

Workstation Benchmarks: Windows 7 vs. Ubuntu Linux

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

  • Workstation Benchmarks: Windows 7 vs. Ubuntu Linux

    Phoronix: Workstation Benchmarks: Windows 7 vs. Ubuntu Linux

    As I alluded to recently, the second round of Windows 7 vs. Linux benchmarks -- with the first round consisting of Is Windows 7 Actually Faster Than Ubuntu 10.04 and Mac OS X vs. Windows 7 vs. Ubuntu benchmarks -- are currently being done atop a Lenovo ThinkPad W510 notebook that is quite popular with business professionals. With the high-end ThinkPad W510 boasting a dual quad-core Intel Core i7 CPU with Hyper-Threading plus a NVIDIA Quadro FX 880M graphics processor, we began this second round of cross-platform benchmarks by running a set of workstation tests. In this article we are mainly looking at the workstation graphics (via SPECViewPerf) performance along with some CPU/disk tests.

    http://www.phoronix.com/vr.php?view=15171

  • #2
    The disk IO bench is the most interesting one. It's the perfect example of why having more throughput is not better; Linux does it faster, but the whole system locks up while doing IO. Windows is slower, but the system stays fluid while the operation is executing.

    No benchmark ever done by Phoronix shows that.

    Comment


    • #3
      Unfair benchmark

      Hi.

      I just registered to say that the article is dead wrong. IOzone needs Cygwin under Windows (it uses mmap() and pthreads), which is an emulation layer. Benchmarks should never be done with emulation of any sort.

      Try running a file system benchmarking software that's native to Windows. Use Wine under Linux and see what happens...

      Cheers,
      Bogdan

      Comment


      • #4
        only 2 syntethic tests.. fail....
        why dont test Enemy Territory ?
        gimp, audacity, openoffice, vlc, compressing stuff?

        Comment


        • #5
          Why always Ubuntu? It'd be nice to see real Enterprise Linux compared to Win 7 results. Red Hat EL and Novell EL.

          Comment


          • #6
            Good article. Possible emulation issues aside affecting performance, I'm not surprised both operating systems were so very close in performance. MS has done a nice job (better job?) engineering this release of their OS. Though I'm mainly a Linux guy, I can say "Good JERB MS!". I've even gone as far as dual booting Win 7. My first MS OS since Win2k.

            Comment


            • #7
              I agree with boobies
              And I'm looking forward to see some AMD benchmarks as well!

              Comment


              • #8
                Not only did Ubuntu 10.04 LTS do a great job over Microsoft Windows 7 with the disk tests, but also when looking at the raw CPU performance of the Intel Core i7 720QM using OpenSSL with RSA 4096-bit encryption the performance
                of Ubuntu Lucid was more than doubled when compared to Windows 7 Professional.
                That doesn't make any sense. In a CPU bound test the OS can't in any way effect the results by such a margin.
                Your test has some major issue, fix it before publishing misleading results

                Comment


                • #9
                  Originally posted by RealNC View Post
                  The disk IO bench is the most interesting one. It's the perfect example of why having more throughput is not better; Linux does it faster, but the whole system locks up while doing IO. Windows is slower, but the system stays fluid while the operation is executing.

                  No benchmark ever done by Phoronix shows that.
                  I think you're wrong. Here's the issue probably and finally, it seems to be fixed soon:

                  http://lkml.org/lkml/2010/8/1/40

                  Comment


                  • #10
                    Originally posted by Enrox View Post
                    That doesn't make any sense. In a CPU bound test the OS can't in any way effect the results by such a margin.
                    Your test has some major issue, fix it before publishing misleading results
                    Why it can't? Search for OS X benchmarks and you'll see Linux also outperformed it in this test.

                    @Love4Boobies

                    Try running a file system benchmarking software that's native to Windows. Use Wine under Linux and see what happens...
                    +1. Btw. I'd like to see what happens.

                    Comment


                    • #11
                      Originally posted by RealNC View Post
                      The disk IO bench is the most interesting one. It's the perfect example of why having more throughput is not better; Linux does it faster, but the whole system locks up while doing IO. Windows is slower, but the system stays fluid while the operation is executing.

                      No benchmark ever done by Phoronix shows that.
                      then you must be doing something wrong

                      for me it's somewhat the opposite:

                      I can work with Windows 7 most of the time but during heavy copying - and especially during antivirus scanning (set to high priority), indexing of files with copernic desktop, google desktop, etc. you simply can't work with it - the delays take MINUTES
                      - this really shouldn't affect productivity

                      compare that to my Gentoo system with 2.6.34 or 2.6.35:
                      even during heaviest transfers, compiling, etc. it's still responsible and I'm NOT using BFS or BFQ - only stock CFS and CFQ - it's not even locking up when doing a desktop search indexing, copying my roughly 800 GB /home partition to another disk (both encrypted), hearing webradio, surfing, etc. etc.
                      - the only problem also taking MINUTES to react is when it's swapping out memory but that's a problem due to heavy swapping when I use all of my memory (not the typical case)

                      Comment


                      • #12
                        Originally posted by RealNC
                        The disk IO bench is the most interesting one. It's the perfect example of why having more throughput is not better; Linux does it faster, but the whole system locks up while doing IO. Windows is slower, but the system stays fluid while the operation is executing.
                        Originally posted by kraftman
                        I think you're wrong. Here's the issue probably and finally, it seems to be fixed soon:

                        http://lkml.org/lkml/2010/8/1/40
                        Both interesting observations. I can't speak much for the situation in Windows since I only use it to play one game (XP). The few times I touch a Vista laptop from some friend I find it lagging with lots of disk I/O. What I can tell for sure is that Linux doesn't behave ideally either. Be it Firefox touching its database or some other program doing something, the whole system is not responsive for as long as the read/writes last. It would be excellent if that commit fixed this issue.

                        Comment


                        • #13
                          Originally posted by yotambien View Post
                          Both interesting observations. I can't speak much for the situation in Windows since I only use it to play one game (XP). The few times I touch a Vista laptop from some friend I find it lagging with lots of disk I/O. What I can tell for sure is that Linux doesn't behave ideally either. Be it Firefox touching its database or some other program doing something, the whole system is not responsive for as long as the read/writes last. It would be excellent if that commit fixed this issue.
                          There's definitely some problem in both. In 32bit XP when system was doing something in the background (antivir or some other activity, maybe refreshing Add/remove programs) it become unresponsive. There's similar behavior, but in 64bit Linux when copying large files. Me and my friend didn't notice this is 32bit Linux and afaik he didn't notice this in his Windows. I hope it will be fixed this time, because it can be very irritating.

                          Comment


                          • #14
                            Originally posted by kraftman View Post
                            Why it can't? Search for OS X benchmarks and you'll see Linux also outperformed it in this test.
                            So the test has issues on OS X as well.
                            In a CPU raw test there's no way the OS can have that effect, as simple as that...
                            To do a comparison that is meaningful you should always test native code... who cares if a test written for Linux and ported with some type of compatibility layer is slower on Windows!
                            What's the point in such a test? Do you want me to write a test on Windows and then port it poorly on Linux to get opposite results? LOL!

                            Comment


                            • #15
                              The problem I mentioned earlier is not the only one. For instance ext4, NTFS and even HFS+ (I see OS X mentioned) are all journaled file systems. Was journaling enabled in these tests? I don't see this information mentioned anywhere.

                              I don't mean to offend the author but perhaps someone more competent should do these benchmarks, i.e., someone with some knowledge of how operating systems actually work. Knowing how to run a program does not count as OS knowledge.

                              Cheers,
                              Bogdan

                              Comment

                              Working...
                              X