Can you use Intel's Turbo boost feature on new 27" iMac?

Discussion in 'iMac' started by alex007, Dec 3, 2011.

  1. alex007 macrumors newbie

    Joined:
    Dec 1, 2011
    #1
    Hello eveyone,
    I was wondering how if possible to use Intel's turbo boost feature/program on a new iMac 27"...? Thank you in advance!
     
  2. The-Pro macrumors 65816

    Joined:
    Dec 2, 2010
    Location:
    Germany
    #2
    Turbo Boost is not something which you can control. Its all done automatically by the processor and system.
    When you have one active thread the system will automatically engage turbo boost. This will overclock one core to the specific clock rate that processor has set (different for every chip). With two active threads turbo boost will once again engage but at a slower maximum clock, usually like 100MHz under the maximum turbo of one thread. This trend continues. When all threads are in use then, if im not mistaken, all cores will be overclocked by 100MHz. If they are not then they will operate at the given speed of the processor, ex. 3.1GHz (for the 3.1GHz i5 in the 27")

    See here:
    http://www.intel.com/content/www/us...ology/turbo-boost/turbo-boost-technology.html
    and here:
    http://en.wikipedia.org/wiki/Intel_Turbo_Boost
    for more information.
     
  3. jvpython macrumors 6502

    Joined:
    Aug 25, 2011
    Location:
    New Zealand
    #3
    Turboboost over clocks the processor a lot more than 100 Mhz. The i7 iMac 27" over clocks from 3.4Ghz to 3.8Ghz and I think there is a similar difference for the i5.
     
  4. The-Pro, Dec 4, 2011
    Last edited: Dec 4, 2011

    The-Pro macrumors 65816

    Joined:
    Dec 2, 2010
    Location:
    Germany
    #4
    "100Mhz under the maximum turbo of one thread"
    To clarify more, it decreases with increments of 100MHz. With one active thread/core it turbos to 3.8 (on the i7, didn't know the exact clock which is why i didn't write that last post). Then when 3 or 4 threads are being utilized it still turbos but 100MHz less. etc.
     
  5. NMF macrumors 6502a

    NMF

    Joined:
    Oct 27, 2011
    #5
    So here's my question. The higher stock 27" has a 3.1 GHz processor, which we know to be the i5-2400. The lower stock 27" has a 2.7 GHz processor, which we know to be the i5-2500S. The interesting thing here is that while the i5-2400 turbos from 3.1 GHz to 3.4 GHz, the i5-2500S turbos from 2.7 GHz to 3.7 GHz. So, when choosing between these two models, would it not be the more intelligent choice to go with the lower-end stock model? If you were to simulate a situation in which all four cores on both chips were being taxes, the 2500S would be 400 MHz behind the 2400. However, when if we were to simulate with only 3 cores, the difference drops to just 100 MHz. With two active cores, the 2500S pulls ahead with a 300 MHz lead (3.6 GHz to 3.3 GHz), and with only one active core the lead is unchanged.

    So basically, of you're running an app that uses only 1 or 2 cores, the 2.7 GHz i5-2500S is actually the faster processor. Don't most modern apps use only 2 cores anyway? Additionally, the 2500S draws 65w of power, compared to the 95w draw of the 2400. Less power means less heat, which means slower fans at low/idle speeds, which means less noise.

    So to recap: the 2.7 GHz chip is more power efficient, quieter, cooler, yet runs at faster speeds in dual or single-core applications. Assuming nothing else is important to you but the CPU (and you have to pick a stock model), the 2.7 GHz model is actually the better buy then, right?
     
  6. leman macrumors 604

    Joined:
    Oct 14, 2008
    #6
    Well, benchmarks show that i5-2500S is actually somehow slower than the i5-2400, it seems that besides the lower base clock it is also more conservative in regards to the over-clocking. But it may end up faster in applications which do not efficiently utilize multiple cores.

    So, if you do lots of CPU-intensive work (like video encoding), your best bet is to pick a model with an i7 CPU. If you need a baseline model, get the i5-2400

    P.S. Geekbench scores sum it up neatly: http://www.primatelabs.ca/blog/2011/05/imac-benchmarks-mid-2011/
     

Share This Page