geekbench

Discussion in 'General Questions' started by xLars, Mar 6, 2007.

  1. xLars

    xLars Bit poster

    Messages:
    3
    How is it possible that XP on Parallels (OSX) geekbenchs much higher than on XP ?

    http://www.geekpatrol.ca/geekbench/

    Does XP runs faster on Parallels than on a PC ? Is it possible ?
     
    Last edited: Mar 6, 2007
  2. constant

    constant Forum Maven

    Messages:
    1,010
    .
    Could it be that windohs in Parallels, benefits in the same way as windoze in Wine?

    Windblows in Wine, runs faster than native windwoes. This is because Linux, which like OSX, is a Unix derivative, and therefore has the ability to pre-execute program code, and store the result in memory. I believe this is known as buffering, and speeds program execution.
    .
     
  3. xLars

    xLars Bit poster

    Messages:
    3
    Is it possible to get an official answer from Parallels development team ? thanks,
     
  4. constant

    constant Forum Maven

    Messages:
    1,010
    .
    Well, one thing you certainly DO NOT want to do, and that's hold your breathe while waiting.

    Parallels' support is absolute crap.

    Did you know that Virtual Box is free http://www.virtualbox.org.

    Also Vmware have a free version at http://www.vmware.com.
    .
     
  5. xLars

    xLars Bit poster

    Messages:
    3
    > Parallels' support is absolute crap.

    :-(

    Yes I know those other products. I am curious about Parallels as XP seems to run very fast on it. Faster than on a PC. It also installs (XP) very fast
     
  6. drval

    drval Pro

    Messages:
    490

    Send your question directly to [email protected].
     
  7. Purplish

    Purplish Forum Maven

    Messages:
    521
    The link you referenced states that Geekbench runs twice as fast in Windows on Parallels on Mac as it does in OS/X.

    To me, that does not seem likely. Geekbench is probably not accounting for the multiple processors correctly in the VM.
     
  8. jfpoole

    jfpoole Bit poster

    Messages:
    1
    Geekbench (both 2006 and 2.0) relies on high-resolution timers that can be inaccurate in virtual machines like Parallels and VMWare. There's an excellent whitepaper on the VMWare website that explains the issues involved; unfortunately I've lost the link to it.

    Anyway, when the timers that Geekbench relies on are inaccurate, you end up with inaccurate results (like Windows XP being 2x faster than Mac OS X on the same hardware).

    P.S., I'm the developer behind Geekbench. I found this post through Google and thought I'd respond.
     
  9. Ynot

    Ynot Pro

    Messages:
    387
    jfpoole, seems you are right. :)
     

Share This Page