RMBP 2.3 vs 2.6/2.7 CPU

Discussion in 'MacBook Pro' started by kimovski, Jun 19, 2012.

  1. kimovski macrumors regular

    Joined:
    Mar 22, 2011
    Location:
    Macau
    #1
    Will I see huge differences between them in terms of performance?
    For:
    1- web,email,office etc... (not heavily stuffs)
    2- Gaming, sound/video editing, graphic design etc...(heavily)

    I mean only the CPU itself.

    cheers :D
     
  2. Bendix macrumors newbie

    Joined:
    Jun 18, 2012
    #2
    No, not really.
     
  3. OSMac, Jun 19, 2012
    Last edited: Jun 19, 2012

    OSMac macrumors 65816

    Joined:
    Jun 14, 2010
    #3
    About 10% at best for gaming and video encoding, trying to decide the same here, it's 3.3 vs 3.6 in turbo mode.

    I think one would seldom notice it.

    The added SSD space would be nice but with USB 3 and TB not sure its needed as an external drive/SSD will run extremely fast. In fact I could see someone coming up with a TB or USB 3 add-on that would mount to the back of the display for adding more drives.

    I'm leaning towards the base model at this time, they are easier to sell down the road as well. I've found people in the used market generally want to spend less.
     
  4. hi-there macrumors member

    Joined:
    Mar 8, 2007
    Location:
    Melbourne
    #4
    For most thing OP mentioned, no noticeable performance improvement from CPU alone. However, the level of L3 cache combined with the added speed will make a difference to gaming and video work. Is that enough to justify a big jump in price tag? Only you can answer that. I personally, don't like to regret later for not going with faster/bigger/etc., and rather pay the price for that piece of mind.
     
  5. brand macrumors 601

    brand

    Joined:
    Oct 3, 2006
    Location:
    127.0.0.1
    #5
    That just sounds atrocious.
     
  6. daleski75 macrumors 65816

    daleski75

    Joined:
    Dec 10, 2008
    Location:
    Northampton, UK
    #6
    SSD would make much more of a perceived impact than a .3 mhz processor speed increase.

    Had SSD's in all my recent laptops and will never go back to a normal hard drive again and will sacrifice cpu speed in order to get a SSD as well for a custom order with Apple (Macbook Air does not count)
     
  7. OSMac macrumors 65816

    Joined:
    Jun 14, 2010
    #7
    Would be if velcro was used :)

    256GB is fine here, everything is on the server.
    If one wants to install bootcamp more space might be useful.
     
  8. kimovski thread starter macrumors regular

    Joined:
    Mar 22, 2011
    Location:
    Macau
    #8
    I'm trying to decide also which one should I get :)
    So the difference is 600usd, how much average i'm paying for the 0.3 speed increase?
     
  9. OSMac macrumors 65816

    Joined:
    Jun 14, 2010
    #9
    It's the upgrade to a 512GB SSD your paying for too ....

    A standalone Intel 240GB SSD is about $300, add the Apple tax its $600 :)
     
  10. Evil Spoonman macrumors 6502

    Joined:
    Jan 21, 2011
    Location:
    California
    #10
    The short answer is no.




    The long answer. There are two primary vectors to discuss.

    1. Applications that execute on a single thread.
    2. Applications that execute on two or more (many) threads.


    1. Single Threaded
    An application that can only take advantage of one virtual core (or thread) is known to be single threaded. There are several reasons for this:
    • There is no reason to add support for additional threads. It would just add needless complexity.
    • The program is poorly written.
    • The problem it is trying to solve is linear and will not benefit from additional threads.

    Most of your applications will consume less than one core's worth of performance at any time. Having multiple cores allows you to run multiple single-threaded programs at once without suffering any slowdown. Many of the more basic programs you use everyday will be single threaded, and it should almost never become an issue.

    In order to increase the performance, one must change the microarchitecture, or increase the clock speed. Intel has decided to use a technology called Turbo Boost to increase the clock speed of the processor based on the thermal overhead and/or number of cores active. This existed with mobile Westmere (Arrandale), but it was unreliable and would not step up bins as often as one would hope. With Sandy Bridge, Intel made Turbo Boost much more consistent and reliable. Most of the single-threaded performance difference between Ivy Bridge and Sandy Bridge is that Ivy Bridge boosts higher more often.


    2. Multi Threaded
    This kind of application can take advantage of however many cores it was programmed to. Raytracers and radiosity solvers such as Mental Ray can take advantage of as many cores as it has available. Other programs like Photoshop, Handbrake, and Starcraft 2 have a limited amount of parallelisation they can perform. Starcraft 2 is infamously limited to two threads, whereas Handbrake seems to run into a limitation in the encoder and maxes around six threads (at least last I used it). Certain parts of Photoshop CS5 and CS6 are well threaded, other parts are often very poorly threaded.

    It is worth looking into the programs you use and figuring out which are most intensive. If you are primarily running a stressful single-threaded program, a case can easily be made for the faster CPU. If your workloads are more multi-threaded, the difference between CPUs shrinks.


    Analysis
    I have attached to this post three charts which rank several Intel processors.

    Cinebench R10 Singlethreaded shows the single threaded performance as measure in Cinebench R10 rendering on only one core. The 2.3GHz current-gen chip is roughly keeping pace with the highest performance SBN chip from last generation. Delta is about 8% vs the high end IVB. This is worst case scenario, there just aren't enough brute force cycles to keep up.

    Cinebench R11 Multithreaded shows the multi-threaded performance as measured in Cinebench R11. The renderer in R11 is substantially improved over R10, and is more representative of the actual performance. As we can see the low-end IVB outpaces the previous highest BTO SBN considerably. The difference between all three IVB chips is fairly small. About 4% for 3615 vs 3720.

    Geekbench Scores paints a more general picture. Geekbench is useful for comparing machines from more of a general synthetic performance standpoint. The low end IVB is the equal of high-end SBN from last gen. Anybody who has a top-of-the-line 2011 SBN MBP can tell you it is wicked fast.


    Processor Legend:
    • i7-3770K - This is a current generation Ivy Bridge desktop part rated at 77 watts. Having this included helps illustrate exactly how close mobile parts are getting to desktop parts these days.
    • i7-3820QM - The 2.7GHz BTO high end part you can spec in the top model MacBook Pros today. This is an Ivy Bridge part, and has 100MHz increased frequency, and 2MB additional cache over the standard high-end part.
    • i7-3720QM - Standard 2.6GHz high-end part in both Retina and non-Retina mid-2012 MBP.
    • i7-3615QM - A slightly slower part, 2.3GHz low-end MBP Ivy Bridge part from this generation.
    • i7-2860QM - The BTO high end Sandy Bridge part from last generation MacBook Pros. This has 8MB of L3 as well.
    • i7-2760QM - High-end Sandy Bridge part for last generation MacBook Pros.
    • i5-520M - Much slower SKU from mobile Westmere. This was in the low end 15" MBP from 2010. It helps frame the bottom of the chart, and illustrates how far we've come since 2010. As well as how much of a performance improvement SBN was.


    NOTE: All of my data here was taken from aggregate internet sources and charted myself in Numbers. It is meant only as a general guide and could have inaccuracy. However, all sanity checks I've performed show that it is quite representative of real-world performance.
     

    Attached Files:

  11. nickbarbs macrumors regular

    Joined:
    Nov 26, 2009
    #11
    You are aware that all chips have the same 6mb cache right?
     
  12. calderone macrumors 68040

    calderone

    Joined:
    Aug 28, 2009
    Location:
    Seattle
    #12
    2.7 has 8 MB of cache.
     
  13. nickbarbs macrumors regular

    Joined:
    Nov 26, 2009
    #13
    ah my bad, standing corrected, I only read the tech specs on Intel's site for 2.3 and 2.6ghz. Honestly for games, even the 8mb of l3 cache won't help, the computer's bottleneck is clearly the GPU, with only 1GB of ram for these huge textures as well as, well, being a mobile GPU. All the CPU in the world won't make up for that, speed, cache, whatever, 2.3@3.3 turbo is enough.

    As an aside I actually just cancelled my order for 2.6/16/512 and rebooked it with 2.3/16/256.

    The price difference couldnt justify, the only reason I wanted the 2.6 was really for the 512 SSD and frankly since its removeable i'm sure in a couple years time i could upgrade if I REALLY need to.

    128gb in the SDHC slot is also an option for holding movies etc.
     
  14. kimovski thread starter macrumors regular

    Joined:
    Mar 22, 2011
    Location:
    Macau
    #14
    Really helpful, thanks for the answers, cheers :D
     
  15. maratus macrumors 6502a

    maratus

    Joined:
    Jun 12, 2009
    Location:
    Canada
    #15
    It's 100$ for processor clock bump
     
  16. jeopardy2311 macrumors regular

    Joined:
    May 25, 2010
    #16

    Thanks a lot for this very useful information. I ordered the 2.6/16/512. I am happy with this order as i couldnt really justify the need to pay the extra money for the 2.7.. Just my opinion..
     
  17. timeslip macrumors member

    Joined:
    May 20, 2007
    #17
    For whatever it's worth.. Difference between the 2.3 and 2.6 are also...

    Intel vPro
    Intel TXT
    Intel VT-d

    I'm not sure if OSX or Fusion really takes advantages of any of these options though.
     
  18. timeslip macrumors member

    Joined:
    May 20, 2007
    #18
    Nm, AnandTech's review says all of them have it. Maybe the screenshot I saw was dated.
     
  19. XiXora macrumors newbie

    XiXora

    Joined:
    Jun 13, 2009
    Location:
    London
    #19
  20. ixodes macrumors 601

    ixodes

    Joined:
    Jan 11, 2012
    Location:
    Pacific Coast, USA
    #20
    Kudos for an exceptionally valuable post.

    For those who took the time to read your well thought out post, they've learned something. Your contribution is stellar.

    For those like myself who are formally trained engineers, working in hi-tech, we can appreciate your efforts.

    Cheers... :)

    ----------

    Your decision was a smart one. I do a lot of very resource intense engineering & design work, although a bit faster the 2.7 is largely a waste of money for the user & a profit maker for the manufacturer. You chose the best of the lot.
     
  21. MH01 macrumors G4

    MH01

    Joined:
    Feb 11, 2008
    #21
    Its like rocking up to a to buy a BMW M3, and being asked to drop some $$$$ to replace the current 19" wheels with 20" wheels
     

Share This Page