5

I have been trying to decide if there will be a substantial difference between the i5 and i7 processor for compiling and running VMs. I expect to run Windows with Visual Studio in a VM as well as a lot of XCode. I don't know how much difference the following benchmark and processor differences will translate to in compiler performance, especially if running them in a VM.

I looked at the Geekbench3 and pulled out the following figures from the averaged charts at Primate labs comparing the two Macbooks against my current desktop.

Benchmark           13” i5 2.8 GHz      13” i7 3.0 GHz      iMac 2.5Ghz i5 21” Mid-2011

64 bit multi-core    6880                7209                 10073

64 bit single-core   3294                3466                  2663

32 bit multi-core    6151                6373                  7103

32 bit single-core   3131                2971                  2466

XCode can compile using multiple cores as can Visual Studio (at least if I boot with Bootcamp, not sure if will work under a VM).

Note that a common mistake is for people to think the i7 in the 13" Macbooks is a 4 core - it's not - both the i5 and i7 in 13" Macbooks are dual-core. The difference is the i7 adds hyperthreading. (The 15" i7 model is quad core.)

The other difference that may be very significant is the 3MB vs 4MB cache.

| improve this question | |
  • 1
    good question, i would like to know also the difference between the 13" i7 dual core and the 15" i7 quad core. i've to bring home the computer every day, the 15"is quite heavy, but I think that there is a big step in performance betweenth i7 2 and i7 4 cores, considering i always keep the simulator on during the compilation – primax79 Dec 4 '14 at 10:49
  • 1
    A major reason for me considering the 13" only is that it will be used for travel including the long Australia-USA flights. If I was thinking of a portable dev machine to be just hauled around locally, I would be going for the top of the range 15" i7 to get the extra GPU. – Andy Dent Dec 4 '14 at 10:56
  • You already have the fact-based part (benchmarks) in the question so as written any answer you get will be primarily opinion-based. Don't know about Apple's return policy in Australia but it might be better to get one of the models, give it a try for a week or two and either keep or exchange it then. – nohillside♦ Dec 4 '14 at 11:14
  • I don't know how much difference those factual differences in the processors will make to the performance of XCode. It's one thing to know about a GeekBench or cache difference. It's another if someone has facts about how that works with XCode. – Andy Dent Dec 4 '14 at 14:02
  • 1
    @primax79 look at the Geekbench scores I link to. The Macbook 15" scores nearly TWICE the score for multi-core 64bit. I have previously worked on very demanding C++ code (1 million line+ CAD software) and parallel compilation made a huge difference to us. If you have big stuff to compile, the multiple cores could make a very big difference and the 15" be a lot faster. I wouldn't worry about the simulator being left running while compiling - I just checked with Activity monitor and whilst idle, the simulator registers no CPU activity. – Andy Dent Dec 4 '14 at 14:27

Browse other questions tagged .