Core i7 upgrade for scientific/matlab computing?

Discussion in 'iMac' started by roninpawnee, Jun 19, 2015.

  1. roninpawnee macrumors newbie

    Joined:
    Jun 9, 2015
    #1
    Considering a retina iMac that I will use to run *single-threaded* scientific computations.. usually code written in MATLAB or Python that has not been written to take advantage of multiple cores.

    Hence I thought the clock speed of a single core is the most important variable and was going to order the 4 GHZ Core i7 version of the retina iMac (instead of a Core i5 or a Mac Pro, all of which have lower clock speeds).

    Does this sound wise?
     
  2. yjchua95, Jun 19, 2015
    Last edited: Jun 19, 2015

    yjchua95 macrumors 604

    Joined:
    Apr 23, 2011
    Location:
    GVA, KUL, MEL (current), ZQN
    #2
    Single core performance between the i5 and i7 in the retina iMac is quite different. The i5 is only 83-84% as fast as the i7.

    http://browser.primatelabs.com/geekbench3/compare/1674278?baseline=2739909

    So yes, by all means, go for i7 if you have the funds to do so.

    PS you should at least get a Fusion Drive setup. If funds allow, go pure SSD. The 3TB drives fail quite easily.
     
  3. jerwin macrumors 65816

    Joined:
    Jun 13, 2015
    #3
    Yopu'd probably get better results if you optimized your code :) Are you going to use an SSD, or a fusion drive?
     
  4. roninpawnee thread starter macrumors newbie

    Joined:
    Jun 9, 2015
    #4
    i was going to get the fusion drive - i know i don't really use more than 256 GB anyway.. so I figured a fusion drive would be as fast as a pure SSD since the "first" 256 GB of a fusion drive is SSD.
     
  5. roninpawnee thread starter macrumors newbie

    Joined:
    Jun 9, 2015
    #5
    actually, I am considering getting a Mac Pro instead of this 5K Retina - I like to plug in my laptop to my desktop screen and this 5K doesn't allow you to do that. I would need to have two big screens on my desk even though I'll never use both at the same time.
     
  6. jerwin macrumors 65816

    Joined:
    Jun 13, 2015
    #6
    first 128 GB.
    https://en.wikipedia.org/wiki/Fusion_Drive

    since you stated single core, I don't suppose you have much need for gpu compute?
     
  7. ThrowerGB macrumors newbie

    Joined:
    Jun 11, 2014
    #7
    Many MATLAB functions are already threaded, and multi-threading your own code for parallel operation is apparently relatively easy. Take a look at http://www.mathworks.com/discovery/matlab-multicore.html
     
  8. yjchua95 macrumors 604

    Joined:
    Apr 23, 2011
    Location:
    GVA, KUL, MEL (current), ZQN
    #8
    You do realize that an nMP is going to be completely overkill, and the i7 5K iMac is more powerful than the baseline quad core nMP.
     
  9. roninpawnee thread starter macrumors newbie

    Joined:
    Jun 9, 2015
    #9
    yes, i do realize that and would much rather get the iMac..

    except that the nMP will allow me to share a monitor with my laptop and a retina iMac's screen can't be used by my laptop. thanks for the sSD tips.

    throwergb, thanks for the matlab tip.
     
  10. lchlch macrumors 6502a

    Joined:
    Mar 12, 2015
    #10
    The difference between the two would be negligible for your purposes.

    Having said that I believe for scientific computation single threaded programs are not ideal. It makes more sense to do the computation on the gpu; which is an order of a mangnitude faster than doing it on the cpu.

    It is not hard to perform gpgpu compute onto your code; you just need the right libraries and a few lines of code.
     
  11. jerwin macrumors 65816

    Joined:
    Jun 13, 2015
    #11
    Speaking of gpu compute, does anybody have figures on double precison FLOPS. I know that the D500 has a marked advantage over the D300 in this regard, but I'm not sure where the m290/m290x/m295x fit
     
  12. lchlch macrumors 6502a

    Joined:
    Mar 12, 2015
    #12
    I can't give you specific numbers. Typically workstation cards have much better double precision performance than the consumer cards.
     
  13. aggri1 macrumors 6502

    Joined:
    Jul 21, 2010
    #13
    If your laptop is a Mac also, target disk mode might be an option with the 5k iMac. Not the most convenient option necessarily, but these things boot so quickly now...
     
  14. jerwin, Jun 23, 2015
    Last edited: Jun 23, 2015

    jerwin macrumors 65816

    Joined:
    Jun 13, 2015
    #14
    there is Gputest 0.7's Julia64 and Julia32

    my results at fullscreen 1080x1920 (I do have a second monitor...) with a m290x

    Julia32: 37509 (fps 624)
    Julia64: 3618 (fps 60)

    meaning double precision is a tenth as slow.

    But in reviewing a D500, the makers of that tool said:

    • FirePro D500: 28514 points, 474 FPS
    • FirePro D500: 9830 points, 163 FPS
    http://www.geeks3d.com/20140425/amd-firepro-d500-mac-pro-late-2013-quick-opengl-test/

    which suggests that the fp64 speed penalty is not as harsh with the D500. IIRC the D700 is similarly efficient, but the D300 is not very good at double precision math.

    (drivers have improved since the D500 review was written)
    Of course, all this is irrelevant if you can't parallelize your code.
     
  15. roninpawnee thread starter macrumors newbie

    Joined:
    Jun 9, 2015
    #15
    This comment got me very excited for a moment before I realized that Target Disk Mode isn't the same as Target Display Mode.. for my setup, Target Disk Mode is a useless feature since everything is in Dropbox and all my computers are instantly in sync. However, the state of my programs is not in sync and it'd be great to just sit down, plug in the laptop and use the same programs..
     

Share This Page