The 'ultimate' test - ish...

Discussion in 'MacBook Air' started by Oli3000, Jul 25, 2011.

  1. Oli3000 macrumors regular

    Joined:
    Apr 20, 2009
    #1
    So - I own the 2010 'ultimate' spec 13" MBA. 2.13/4/256 with the 320m! On saturday, in a moment of madness, I bought the new 'ultimate' 2011 MBA. 1.8 i7/4/256 with the HD3000.

    The whole point of this was we all knew that the new model was '2x' faster for most processes, and just much snappier in general. But I wanted to know the real term meaning of this for things I do, before selling my 2010 MBA for sure.

    This 8 minute video on youtube: http://www.youtube.com/watch?v=XSGBVzeBUbk

    Fan speed rpm (start, end)
    2010 = 2000, 2500
    2011 = 2000, 2000

    Temps degrees celcius (start, end)
    2010 = 35, ~75
    2011 = 39, ~75

    CPU Usage (average)
    2010 = 50-60%
    2011 = 15-20%

    This was actually not what I was expecting - although the difference in cpu usage is amazing! So next I tried a 1.4gb HD .mkv file, playing in VLC player.

    Fan speed rpm (start, end)
    2010 = 2000, 2000
    2011 = 2000, 2000

    Temps degrees celcius (start, end)
    2010 = 45, 65
    2011 = 55, 65

    CPU Usage (average)
    2010 = 30-40%
    2011 = 5-10%

    The higher start temperatures are more common in this test - the lower ones in the first test were attributable to it being straight away after boot. It is also worthy to note that these tests were done in a room probably little over 20 degrees c - good old cold British summer!

    So, at this point, I was more than a little confused! We all know that the 2010 model is meant to go spinning up like mad with streaming HD content etc... So, I tried elsewhere...

    So, I tried another website well known to us Brits - BBC iPlayer - Flash based HD tv content.

    Within 2 minutes of watching a show on here, the 2010 model was at 80 degrees C, with a fan speed of 3500-4000rpm. The 2011 model happily chugged along at, again 80 degrees C, but the fan never budged from 1999rpm. The cpu usage again was very similar to that described above, too; 60-70% for the 2010, and 15-20% for the 2011.

    In general the cpu on the 2011 can be running at near 80 degrees celcius before the fans kick in, at all! The 2010 model is spinning away at anything above 60, normally. The standard operating temperatures for the two seems to be fairly similar, though, and the % of CPU used has a clear victory for the 2011 model!



    Next, gaming! So far I have done two tests, both side by side.

    First:

    SC2. Using the default 'medium' settings (see below), I proceeded to play a 1v1 macro game on each. The base FPS was around 35 for both, in a big battle (particularly with lots of cloaked units), the 2011 faired slightly better, dropping to maybe 11 at its lowest (5 for the 2010). Both were playable, although the 2010 model possibly stuttered more often. This was a dead heat!

    Resolution: 1440x900 native
    Texture: Medium
    Shaders: Medium
    Lighting: Low
    Shadows: Medium
    Terrain: Medium
    Reflections: Off
    Effects: Medium
    Post-Processing: Medium
    Physics: Low
    Models: High
    Unit Portraits: 3D
    Movies: High
    V Sync: Off


    Next, Portal 2. I tested this both online in co-op, and in single player, using the same level each time in both experiments. The settings here were again identical, and using a high setting (see below).

    Resolution: 1440x900 native
    AA: 2xMSAA
    Filtering Mode: Anisotropic 4x
    Wait for V-Sync: Enabled (Double Buffered)
    Multicore Rendering: Enabled
    Shader: High
    Effects: High
    Textures: High

    They both averaged around 20-25fps in co-op, but with the robust source engine meant this is very very playable! The 2011 model dropped to about 12fps in environments with lots going on (goo & light bridges & portals), and the 2010 model coped better at around 15fps. But again, both are very playable.

    In single player I noticed some interesting observations, e.g. if you stood and stared at a energy shield or something with lots of shaders, the 2011 model averaged 19fps, the 2010, 29fps. If you stood and stared at something with less shaders, they would both average 29fps. When moving around they would both perform roughly the same between 20-25fps. The 2010 model noticeably 'stuttered' more often though, especially when entering new rooms. The 2011 model coped with this much better, potentially making it the more playable of the two!

    During playing either game both had fans going at between 6000-6500rpm, and temperatures of 80-90 degrees C. Unfortunately this is quite loud, but that is always going to be unavoidable with this form factor!


    Other points of interest:

    - The 2011 MBA speakers sound a little 'flatter' than my 2010. I put them side by side and swapped between them for sections of songs. Still good, but perhaps a bit less body. Also, annoyingly, one on my 2011 buzzes, whereas the 2010 doesn't. It may be going back because of this!

    - SC2 seems to have some sort of colour issue with the HD3000. Now I'm not sure why this might be, and I've not tried reinstalling, although I have tried repair etc. but the first image is the 2010 MBA, the second the 2011. Look at the difference in the text. This extends to unit health bars etc. but the overall colour of the game is the reason - explanations? Who knows!

    http://imageshack.us/photo/my-images/64/screenshot20110725at213.png/
    http://imageshack.us/photo/my-images/713/screenshot20110725at213.png/

    - Both of mine have LG screens in, but the new one does appear to be marginally brighter - whether that is linked to Lion, I'm not sure, but this is a MINUTE difference!

    - The fan is quieter on the 2011 model, at 2000rpm the 2010 model can be heard ever so slightly from normal working range (desk typing etc.), the 2011 is deadly silent.

    - In a Skype video call the 2010 model runs at roughly 4000rpm, with at least 60% CPU usage. The 2011 model however copes fine, remaining at 2000rpm and burning around 30%! Much better!

    - The 2011 MBA has a NICER keyboard. It feels more like the pro keyboard with big spongey keys. This is unlike the 2010 model that feels like you are going to break the key every time you press it (imo).

    - The fan definitely kicks in later - if the 2011 is running at 80 DC, the fan can still be at 2000rpm - no panic. The 2010 model will likely be at 4000+ at this point to just keep it at bay!

    I have not currently looked at battery life - but may test that out tomorrow. I have not perceived any difference in general use though! The same goes for boot times. The i7 is definitely the snappier of the two machines though!


    To do list:

    - Tomorrow I am going to encode the .mkv file above with handbrake to something different. As we all know, the i7 will be much faster, but I thought I had better put this in writing!

    - I may also xbench both laptops - I am aware this is 'outdated' - but hey, running the same test on both can do no harm!



    So, my current feeling. They are both great machines!! The 2011 model though has to be the clear victor. It is punching above it's weight for gaming, whilst also delivering much faster cpu intensive tasks, and coping with the all important youtube/flash related tasks with far greater ease! The backlit keyboard is nice, although not essential, but the feel of the keys is much nicer as I said before! If you have any questions do not hesitate to ask, and feel free to tear this apart as well - I have never done anything like this before so I have no idea how to track 'average fps' etc. etc. - but if you want me to redo these tests with that, then I shall! I will not be testing WoW as I do not own it, but I could test a few other Valve/Steam games if people are interested.

    I think it can be safely concluded that this laptop moves forward in all areas, apart from those where it stands still. IT DOES NOT GO BACKWARDS IN ANYTHING!


    L&P
    Oli
     
  2. jpi macrumors newbie

    Joined:
    May 6, 2011
    #2
    thanks man. I really liked the review. basically everything addressed that I wanted to know, that is, quietness respectively rpm of the fans. I have a windows computer right now and it drives me nuts, because it is soo freaking loud :mad: so that is definitely a point for me. also like the real life test with games and fps. probably not going to play any but good to know that all this "oh the HD3000 is crap and a huge step backwards" is not all that true.
     
  3. Mandork macrumors member

    Joined:
    May 8, 2011
    #3
    Great comparison tests!
    Can you leave your computers unplugged when doing the next tests, and leave them like that (with caffeine) to see how the battery life is?
     
  4. theSeb macrumors 604

    theSeb

    Joined:
    Aug 10, 2010
    Location:
    Poole, England
    #4
  5. jpi macrumors newbie

    Joined:
    May 6, 2011
    #5
    nice. I guess no need to worry about heat then. good to know. maybe I'll go for the i7 then...
     
  6. Obscurelight macrumors 6502

    Joined:
    Jul 22, 2011
    Location:
    NYC

Share This Page