2017 Early Geekbench Scores seem strange...

Discussion in 'MacBook' started by psymac, Jun 9, 2017.

  1. psymac, Jun 9, 2017
    Last edited: Jun 9, 2017

    psymac macrumors 6502

    Joined:
    Jul 17, 2002
    #1
    Unless I'm looking at it wrong, the MB 2017 m3 and i5 show about the same recent (last 10 or so) Geekbench scores, 3560 for single core, and about 6800 for multicore. So, no real advantage to pay extra for the CPU upgrade then?

    BTW, the 2016 m3 is about 2800/5500, m5 is about 3400/6400, and m7 is about 3100/6600.

    No recent i7 Geekbench scores that I could find, hopefully that is much faster and worth the premium price.

    Still wondering 1) why the 2016 m5 single core is the fastest for 2016, 2) why the 2017 m3 is the faster than the i5.

    Seems then the best value is actually the base 2017 base m3, unless you really need 512SSD and/or 16gb RAM.

    Geekbench

    2017 m3

    http://browser.geekbench.com/v4/cpu/search?utf8=✓&q=MacBook10,1+m3

    2017 i5

    http://browser.geekbench.com/v4/cpu/search?utf8=✓&q=MacBook10,1+i5
     
  2. Eric8199 macrumors 6502a

    Joined:
    Feb 27, 2009
    #2
    Not that this answers your major question, but the m3 is still classified as an m3. The m5 has become the i5 and the m7 has become the i7.
     
  3. psymac thread starter macrumors 6502

    Joined:
    Jul 17, 2002
    #3
    Yes, I see that now on Apple's site, wondering if that has any real meaning from Intel not to brand the m3 as i3.
     
  4. bill-p macrumors 65832

    Joined:
    Jul 23, 2011
    #4
    My m7 scores are higher than what you have:
    https://browser.geekbench.com/v4/cpu/3066613

    ^ just casually ran this while having my Linux VM running a prime number generator in the background.

    Don't rely on Geekbench. It's a "ballpark figure" rather than an absolute metric.

    Everything plays into Geekbench: the memory, the SSD, etc...

    Compare the results of my score versus the m3 and you'll see in some tests that the m3 is able to achieve 40-50% faster transfer speed from the upgraded SSD that allows it to get a higher score.

    Also the CPU varies its speed so it's not clear what performance level you'll be able to benchmark the system at.

    Last but not least, different MacOS build as well. The new MacBook runs a newer build. That could very well play into the performance improvement due to the above reasons.
     
  5. Appleaker macrumors 68000

    Joined:
    Jun 13, 2016
    #5
    Because the m3 would mean an i3 and that doesn't indicate power to people like the i5 and i7 does, which was part of the reason for rebranding.

    The m3 is now aimed towards 2 in 1s and tablets, that's where you'll find it more now on the PC side.

    The price for the m3 and i5 are actually the same, at least for standard Intel pricing. So low gains wouldn't be that surprising, however the i7 should perform a lot better as it has a significant premium.
    The pricing is a reason why I hoped for them rid of the m3 and have i5 and i7 like most other Macs. This mess seems to have caused confusion to many.
     
  6. psymac thread starter macrumors 6502

    Joined:
    Jul 17, 2002
    #6
    Confusing then that the 2016 m5 performs more than the m7, while the 2017 m3 performs much better (so far atsGeekbench scores go) than the i5. Will be very interesting to see the i7 Geekbench scores, hopefully in about a week when they start shipping.

    If this holds, the base 2017 MB m3 model looks like a great bargain, at least over the i5.
     
  7. psymac thread starter macrumors 6502

    Joined:
    Jul 17, 2002
    #7
    Corrected, thanks.
     
  8. Appleaker macrumors 68000

    Joined:
    Jun 13, 2016
    #8
    I think it may be to do with turbo boost, as I believe the 5W processors rely on it a lot more than others. The m3 may need to turbo boost more, hence the higher score while the i5/i7 can manage some of the tasks without turbo boost but naturally it is slower.

    Therefore, the geekbench results wouldn't be the best measurement of performance. That also explains why many of the results are inconsistent with each other, as is the case with previous gen Y-series chips. Although I would imagine the performance gain isn't actually that large.
    Another example of where a geekbench result doesn't show the actual performance of a device is with the iPhone 7 - sometimes it uses the high efficiency cores leading to a much lower result.

    I just can't believe they're at this level already, it seems like a massive improvement from before, especially for the m3. Now I'm just hoping for big leaps in integrated graphics.
     
  9. psymac thread starter macrumors 6502

    Joined:
    Jul 17, 2002
  10. waigx, Jun 14, 2017
    Last edited: Jun 14, 2017

    waigx macrumors member

    waigx

    Joined:
    Jan 21, 2014
    #10
    I second to this, but I don't see any major difference between m3/i5Y/i7Y, they all ultra low voltage processors with the same caches, same maximum 16Gb ram support, etc. The only difference is the clock speed.
    https://ark.intel.com/products/97538/Intel-Core-m3-7Y32-Processor-4M-Cache-up-to-3_00-GHz
    https://ark.intel.com/products/95452/Intel-Core-i5-7Y54-Processor-4M-Cache-up-to-3_20-GHz
    https://ark.intel.com/products/95441/Intel-Core-i7-7Y75-Processor-4M-Cache-up-to-3_60-GHz

    A rather moderate performance difference is expected. I would put my money on RAM (or storage if needed) for a better experience. The MacBook is only made for portable I won't use it do any CPU-intensive stuff.
     
  11. Cloudify macrumors newbie

    Joined:
    Jun 5, 2017
    #11
    You can only squeeze so much blood from a stone... and 14nm is that stone.
     

Share This Page