How much faster is the "Nehalem" 2.66 vs. the original MP?

Discussion in 'Mac Pro' started by Chupa Chupa, Mar 7, 2009.

  1. Chupa Chupa macrumors G5

    Chupa Chupa

    Joined:
    Jul 16, 2002
    #1
    I have an original dual dual-core (2 x 2.66) MP. How much faster is the single 2.66 quad-core "nehalem" MP that just came out with apps like Motion. Anyone have any benchmarks?
     
  2. drlunanerd macrumors 65816

    drlunanerd

    Joined:
    Feb 14, 2004
    Location:
    UK
    #2
    Probably not a huge gain with Motion, given that it's GPU limited.
    Check out Apple's benchmarks for Final Cut Studio 2 HDV rendering - only 1.1x faster for the top end 8 core machines.

    A graphics card upgrade may give you more significant performance gains. Word on the street is that the new ATI Radeon 4870 card may work in the first gen Mac Pro.
     
  3. Tallest Skil macrumors P6

    Tallest Skil

    Joined:
    Aug 13, 2006
    Location:
    1 Geostationary Tower Plaza
    #3
    We don't really know, as no one has theirs yet.
     
  4. Tesselator macrumors 601

    Tesselator

    Joined:
    Jan 9, 2008
    Location:
    Japan
    #4
    With apps like Motion? Depends on what kinda video you're doing. The balance between the 3D functions used and the amount of video data it's processing is kinda critical to a test like this.

    Here's some guesses:

    If it's a you-tube video (SD) I guess 1.1 to 1.2 times.
    If it's a 720p video I guess 1.3 to 1.4 times.
    If it's a 1080p video I bet it's 1.5 to 1.7 times faster.(what Apple themselves benchmarked)
    After that the proformance curve levels off flat.

    For the first two you probably won't notice any difference in the GUI speed. For the 1080p test you may begin feel the CPU getting hit on the older system a little more often than on the newer system.

    All of this speculation is also considering that the storage i/o subsystem is a little faster on the newer machines and also that the newer one accepts faster video cards than the older 2006 ones.

    Unfortunately the usage profiles dramatically on the side of user operation. You probably spend 100 minutes manipulating the GUI gadgets and setting options for every 5 minutes that the processor is being hit hard enough to make a difference. It's for this reason that clock speed does matter a lot on a workstation.

    That's comparing it to the 2006 Mac Pro though (Like Apple did on their website). If you compare it to last year's 2008 machines with the already faster i/o subsystems and the faster video cards available, then I would guess it profiles more like:

    1.0 times
    1.1 times
    1.13 times
    Flat.

    respectively for the three data sizes listed above.

    I hope that helps. We'll know more when people start receiving delivery of the new units and tests like this actually are carried out - hopefully not by people just interested in justifying their purchase or fanboying for Apple. :D


    .
     
  5. bozz2006 macrumors 68030

    bozz2006

    Joined:
    Aug 24, 2007
    Location:
    Minnesota
  6. drlunanerd macrumors 65816

    drlunanerd

    Joined:
    Feb 14, 2004
    Location:
    UK
    #6
    Apple are benchmarking to the 2008 Mac Pro (and the Power Mac G5 Quad).
     
  7. jjahshik32 macrumors 603

    Joined:
    Sep 4, 2006

Share This Page