11" i7 vs i5 Test(s)

Discussion in 'MacBook Air' started by Jazwire, Jun 26, 2013.

  1. Jazwire, Jun 26, 2013
    Last edited: Jun 26, 2013

    macrumors 6502a

    Jazwire

    Joined:
    Jun 20, 2009
    Location:
    127.0.0.1
    #1
    Streaming a HD Netflix (The Lorax) movie test.
    ====================
    Right now I have my MBA 11" i7/8/256 vs my wife's 11" i5/4/128.


    I made sure every program was shut down, checked activity monitors to make sure they were exactly the same before I started. Turned off Time Machine.

    Both charged to 100% Let them even charge for 30 min after they reached 100% to make sure they were maximum charged.

    Both set to 8 brightness ticks and 1 keyboard brightness tick.

    I am streaming both in full screen using Safari simultaneously side by side.

    Results as I have them below:
    ===================================

    22 min into movie:

    i5 - 92%
    i7 - 97%

    i7 is actually cooler on the bottom, neither are really warm, but the i7 heat is actually almost undetectable. Was not expecting these results so far.

    (Double checked activity monitor, nothing different running on the i5, no fans on either machine)

    **
    45 min into movie:
    i5 - 70%
    i7 - 84%

    Here is something interesting I observed, higher cpu usage on the i5 for Safari and silverlight (netflix player plugin.)
    i5 is definitely "warm"
    i7 is very slightly "warm"

    If I were to give a value 1-10 on heat/warmth.
    i5 - 4.5
    i7 - 3

    **
    70 min into movie:
    i5 - 54%
    i7 - 76%

    Heat is the same as earlier with the i5 actually running a little warmer.
    No fans running on either machine at any point thus far either.

    **
    End of movie 1hour 25 min:

    i5 - 42% (1:04 Remaining)
    i7 - 70% ( 3:24 Remaining)

    ---------------------------------------------------------
    Ending thoughts: Was totally blown away, I was expecting the i7 do do slightly worse than the i5 for both heat and battery.


    Possible Conclusion: I did check activity monitor on both machines around the 45 minute mark and watched it for a min or two.
    I noticed the i5 cpu load mainly between 45-70 it bounced around a lot in that range.
    The i7 cpu load stayed in a much tighter range of 39-45. Did not bounce around like the i5 did.

    The cpu load was nearly completely on both machines Safari & Silverlight (Plug-in), the i5 wasn't using CPU cycles on anything else.

    So my thoughts are in this test the i7 just handled the job more efficiently, the i5 was throttling up and down a lot and that caused heat and greater battery loss, the i7 was able to maintain a consistant threshold and didn't suffer from having to "turbo boost" as much because the base speed was sufficient.

    Thoughts??


    (More tests coming, Next, League of Legends (Game) heat, battery & FPS test, side by side... tomorrow night. )
     
  2. Jazwire, Jun 26, 2013
    Last edited: Jun 27, 2013

    thread starter macrumors 6502a

    Jazwire

    Joined:
    Jun 20, 2009
    Location:
    127.0.0.1
    #2
    ** League of Legends Test**

    Setting on both machines.
    Resolution - 1368x768
    Character Quality - High
    Effect Quality - Medium
    Environment Quality - High
    Shadows - Off

    i5 results:
    Avg FPS:41
    FPS Range:37-50
    Heat: Warm, maybe slightly less warm than i7 , fans running fairly loud

    i7 results:
    Avg FPS: 48
    FPS Range:44-58
    Heat: Warm - No Fans (Actually ran i7 test twice, because thought maybe I didn't hear fan on 1st run.)

    I did not test batter drain on this test. Ill save that for the next text. I just wanted to compared performance and heat.

    I was surprised the i5 fans spun up and the i7 did not. Heat was about even, maybe slightly cooler on i5, but it had the fans running where the i7 did not.
    ***********************************************************


    More test tomorrow. (General Usage Test) *Details later* (Sorry tired tonight)
     
  3. Kevbodian, Jun 26, 2013
    Last edited: Jun 26, 2013

    macrumors member

    Joined:
    Dec 23, 2005
    #3
    Nice Jazwire. Probably not the case, but I wonder if the 4GB device had to use swap for the stream more often than the 8GB device... may explain the extra CPU usage.

    Playing a video is a matter of loading the video into memory and decoding whereas streaming is using wifi, downloading, placing portions into memory, decoding, rinse repeat.

    Can you do a CSGO test? I believe the game is natively supported on OS X. I know it may be hard to do a similar test as it's different every time.
     
  4. macrumors 65816

    Joined:
    May 4, 2012
    #4
    At the end of all these tests the conclusion is, "So what?" and "Who cares?".
     
  5. macrumors 65816

    Joined:
    Dec 14, 2010
    Location:
    UK
    #5
    You dont find it useful, but others do.

    Unexpected results, but truely positive...............for me anyways! ;)
     
  6. macrumors newbie

    Joined:
    Jun 23, 2013
    #6
    This

    ----------

    In that case I still think i7 is going to come out on top.

    I'm fairly surprised by the battery results. One would think video decoding is a piece of cake for both current gen i5 and i7, and the battery usage shouldn't really differ much.
    However the base clock speed is pretty low for the i5. 1.3 is probably sufficient for web surfing and PDF/office work, anything higher than that would need turbo boost. If I want to squeeze out extra Battery life, utilities like ThrottleStop or any equivalents in OS X might help.
     
  7. macrumors member

    Joined:
    Dec 23, 2005
    #7
    I bought an i7/8/128 and it was delivered the day after I went to Japan for 2 weeks (still there now). For the last week or so I've been thinking long and hard whether I should return the i7 and get an i5/8/256. I'm beginning to think I'm alright with the i7.

    For me battery life is important as most of the work I do on the laptop will be while on battery. It's just nice to be able to play a game or two once in a while when I'm away from home (I have a nice gaming PC..). Lol, I love when people like kodeman come in and say nothing productive... or come in with "if you want to game, why'd you buy an air"? haha
     
  8. macrumors 6502a

    mayuka

    Joined:
    Feb 15, 2009
    #8
    @Jazwire: Could you test idling performance? :confused: At least, that's what's most important for students that are writing during courses or in the library.
     
  9. macrumors member

    Joined:
    Dec 23, 2005
    #9
    Yes, please test sleep while you're at it ;)
     
  10. macrumors 65816

    HiDEF

    Joined:
    Jun 23, 2010
    Location:
    Miami, FL
  11. macrumors regular

    paeza

    Joined:
    Oct 23, 2006
    Location:
    Canberra, Australia
    #11
    Thank you so much.
    Please keep posting more tests.
     
  12. macrumors regular

    Joined:
    Jun 17, 2013
    #12
    Before jumping on to the bandwagon of how awesome i7 is - let it go through some more practical real world usage of:

    1. Heavy surfing - atleast 3 different browsers and 12 different tabs in them
    2. Heavy word processing - 20 files open at once
    3. Heavy PDF processing - 10 files open
    4. 3 hour video - 4 hour audio
    5. Flash videos (the xxx sites) - let them run in background :)

    & then do the same on i7

    I am not a graphic designer & video encoding isn't something I will ever do ..: max - I am going to use adobe products on it like Lightroom - photoshop & some other word/PDF programs & it's mostly will be a business laptop (NO gaming) - so I think my 11" 15/8/512 should be good enough to handle it all nice and cool ...
     
  13. macrumors regular

    Joined:
    Jan 31, 2011
    #13
    Is this really practical? How many people have 36 tabs or 20 word docs open simultaneously?
     
  14. macrumors regular

    Joined:
    Jun 21, 2013
    #14
    +1 that would drive me crazy having that many things open at once
     
  15. macrumors 65816

    musika

    Joined:
    Sep 2, 2010
    Location:
    New York
    #15
    Sometimes when johnjey is viewing his 36 different "xxx sites", he likes to knock out a few word documents and a spreadsheet or two. Is that really so wrong!?
     
  16. macrumors 6502

    Joined:
    May 27, 2013
    #16
    Unless the two models were identical apart from the processor change, you can't draw any conclusions from this test. It could be the extra RAM/SSD benefitted the i7, and there also could be software conflicts that the OP isn't listing. Not really a true test at all.
     
  17. macrumors 6502

    Joined:
    Dec 1, 2008
    #17
    Your results does not make sense. Obviously the i5 was doing something in the background, spotlight indexing perhaps? You certainly won't have better battery life with the i7 and definitely not that much.
     
  18. Guest

    falconeight

    Joined:
    Apr 6, 2010
    #18
    Streaming a netflix movie isn't a CPU intensive test. Its an easier task for a more powerful machine.
     
  19. macrumors member

    Joined:
    Dec 23, 2005
    #19
    John, don't turn this into another of your flame threads. The point of this thread is to provide insight into the differences in performance between the i5 and i7. It's really just to let people make up their own minds as to what's important for their computer. It's not about bandwagons or super thorough benchmarks -- none of 1-5 actually mean anything.

    Jazwire is using his personal time to do some basic comparisons so that the rest of us have some idea.

    ----------

    It wasn't really supposed to be CPU intensive. It was just a simple way to run down the batteries of each machine.

    For the sake of the world, maybe thread title should be changed from "test" to "ramblings" so that people don't take it so f'n seriously. :p
     
  20. thread starter macrumors 6502a

    Jazwire

    Joined:
    Jun 20, 2009
    Location:
    127.0.0.1
    #20
    Sure I can do that.
    It will have to be later, wife took her i5 to work. :/
     
  21. macrumors member

    Joined:
    Dec 23, 2005
    #21
    Lol, the proof is in the pudding.

    ----------

    Lol no don't! Haha I was being sarcastic to the poster above. :) Wait... I see what you did..
     
  22. thread starter macrumors 6502a

    Jazwire

    Joined:
    Jun 20, 2009
    Location:
    127.0.0.1
    #22
    I should be able to accommodate some of these tests to some degree.
    I'll come up with a stress test that I run on both machines side by side.
    Once again it will have to be several hours from now, wife has her i5 at work.

    ----------

    I'd agree with you if the tests were reversed.

    The i7 SSD should and likely does actually draw more power. Each machine had 0 page outs, and were only using around 1 GB of active ram.

    Also there was NO software conflict, I opened up activity monitor an triple checked nothing was running that was not running on either machine. I looked before I started the test and look while I was running the tests.
     
  23. thread starter macrumors 6502a

    Jazwire

    Joined:
    Jun 20, 2009
    Location:
    127.0.0.1
    #23
    It was not, i checked Activity Monitor prior to starting the test nothing else was running that was not on both machines. I checked the Activity Monitor at 45 min into the movie again. Nothing different was running, (though like i said earlier, the i5 was working harder on the safari and silverlight process than the i7).

    It does not make sense which is why I checked the Activity monitor 45 min into the movie.

    Also my wife got her i5 the day after launch, she hardly has anything on it about 25 gb, everything on her end was indexed by spotlight last week and I had time machine turned off on both units.
     
  24. thread starter macrumors 6502a

    Jazwire

    Joined:
    Jun 20, 2009
    Location:
    127.0.0.1
    #24
    Never was meant to be a CPU intensive test, more a real world test.
    People do stream a lot, this was more for heat and battery.

    And if you read my entire post, I came to the same conclusion , basically you did. i7 is the more powerful machine and did not have to "turbo boost" as much.

    I do plan on doing a more CPU intensive test, which logically the i7 should not win in battery time and most likely heat. (Of course I didn't expect it to win this test either)

    However, what ever I do to do that test, its far more likely people are going to stream video than run their cpu at 100% until it dies.

    I am doing some practical and willing to do some impractical tests on both machines. But I am just doing these on the side on my time. And I am doing a thorough job making sure the conditions are exactly the same. Heck, I even ran them side by side at the same time, so that they would have matching room temperatures and same amount of air circulation in the room.

    Do not shoot the messenger with the results.
     
  25. macrumors newbie

    Joined:
    Jun 23, 2013
    #25
    I don't think you need i7 to run the games. From what I read about gaming performances, i5 or i7 doesn't really make much of a difference. I'll expect i5 to give more loading time but frame rate-wise you probably won't be able to distinguish between them.

    What I concluded from all these i5/i7 hassle is what some people have always been saying: if you're not absolutely sure what you need the i7 for, you probably don't need it.
     

Share This Page