When Steve Jobs announced that the MacBook Air would have a 5 hour battery life, I laughed. Short of Apple sticking a hydrogen fuel cell in the Air there was simply no way you were going to get 5 hours out of this thing under any sort of heavy usage. What shocked me even more is how few questioned the 5 hour figure. Have we all forgotten about optimistic battery life ratings? I like to give Apple the benefit of the doubt and not call out RDF whenever a new product with exciting features launches but seriously folks - 5 hours?

We're still hard at work on the MacBook Air review, but this data was too important not to share right away. By now you've heard that under "normal" usage, the Air can only manage about 2.5 hours of battery life. That's true and at the same time it isn't. I touch on part of this in the review, but the MacBook Air isn't designed to be your work notebook - no ultra portable is. It's a notebook for a writer, for someone taking notes in class; it's a second notebook, or a third system. As such, the typical usage model can be very different than your standard notebook.

At the same time, most of the people considering the Air aren't looking at it as a second notebook or a third system, they are looking at it as their primary notebook (and will be left sorely disappointed or frustrated after a while). The end result is that you need a handful of battery life tests to really cover all bases. And that's what I've done.

I scripted together three tests: wireless web browsing, DVD playback and a multitasking workload.

The wireless web browsing test uses the 802.11n connection to browse a series of 20 web pages varying in size, spending 20 seconds on each page (I timed how long it takes me to read a page on Digg and came up with 36 seconds; I standardized on 20 seconds for the test to make things a little more stressful). The test continues to loop all while playing MP3s in iTunes. This test is designed to simulate the intended usage of the MacBook Air: something you can carry around with you to class, work, on the train, etc... to comfortably and quickly browse the web, take notes and generally be productive all while listening to music. It's like a big iPhone...without the phone part.

Despite the lack of an internal optical drive (I've never been so tempted to use the f-bomb in a review before), the MacBook Air will play DVDs. You just kind of need to have them on your hard drive and I'll just assume you have the original disc safe at home. This test is simple: I play Blood Diamondin a loop until the battery runs out.

The final test is the multitasking workload, and this test isn't really that intensive for a normal system but thanks to the terribly slow iPod hard drive in the Air - it's quite stressful. For this benchmark I'm downloading 10GB worth of files from the net (constant writes to the drive), browsing the web (same test as the first one) and watching the first two episodes of Firefly encoded in a 480p XviD format (Quicktime is set to loop the content until the system dies). There's nothing too extreme about this workload, I could definitely come up with worse - but this would be light to moderate usage on a MacBook or a MacBook Pro, and it'll show how the Air stacks up.

The system was set to never shut off the display and never go to sleep, although the hard drive was allowed to spin down when possible. The display brightness was set at 9 blocks (just over 50%), which I felt was comfortable for both day and night viewing. The system itself was a standard MacBook Air with an upgrade to a 1.8GHz processor, but used the regular PATA hard drive - not the SSD (I'll touch on SSD performance once I get my drive).

Ready for the results?

Apple's 5 hour claim is laughable but not as much as I expected. If I wanted to I suspect I could hit 5 hours by making the web browsing test less stressful, but my focus was on real world usage scenarios, not proving Apple correct. Regardless, 4 hours and 16 minutes doing what I consider to be the intended usage model of the Air is respectable. It's not great, but it's not terrible either.

The DVD playback test is also decent, you can watch a full length movie on the Air (even The Godfather). But watching two back-to-back isn't possible (although I do think you could squeeze two viewings of Rambo in there and frighten your fellow airline passengers).

And then we have the multitasking scenario. At just under two and a half hours, this isn't your do-everything notebook. That should've been obvious from the get-go with the whole lack of an optical drive, 2GB memory limit and no built-in Ethernet port.

I'll save more thoughts and discussion for the full review but this data was too important not to share.

On a side note, in doing the DVD playback testing on the MacBook Air I realized something I hadn't previously about Leopard. It looks like Apple finally fixed the horrible de-interlacing problems with its previous DVD player application. There are now actual de-interlacing controls and image quality seems vastly improved; I may have to run HQV to see what sort of improvement there is over Tiger. What took me so long to notice this about Leopard? Well it seems like Apple has finally fixed issues with DVD playback, just when everyone is starting to move to HD-DVD and Blu-ray. Great.

Be sure to read our other MacBook Air coverage:

Inside Apple's MacBook Air

The MacBook Air Ethernet Dongle: Performance Explored

Comments Locked


View All Comments

  • macbookair - Friday, May 9, 2008 - link

    I recently bought the macbook air- my first mac- not expecting much from it but unable to resist the ascetics.

    I've found with MY normal usage patterns, which are probably fairly standard for uni students/office workers; (taking notes/ writing and doing research- so browsing and downloading TEXTS from the internet) as well as normal email/networking activity- that the battery certainly does last 5 hours. In fact even after it has 'apparently' reached the last minute of use it continues for another 1/2 hour or so (at least that's as it's as far as I've pushed it).

    Plus it fits comfortably in my handbag and is lovely and light (so I can easily work on the bus/train and in the park). I don't really need to play DVDs etc when out of the house... so I'm not disappointed!

    It's only flaw so far, for my purposes, is that it's so damn pretty that I'm very precious about it!

  • macbookair - Friday, May 9, 2008 - link

    ps... thats without "better battery life" switched on or dimmed screen etc...
  • mynest - Sunday, October 24, 2010 - link

    We ran a typical battery of tests(you can go to http://laptopsbattery.co.uk/blog/ to get more infomation), including games, applications, and a couple other areas. Time constraints limited the amount of testing we could get done before the NDA lift (and it's already a few hours late), but we will be providing a follow-up article with further testing and information. That said, if there are any specific areas that you'd like to see tested, feel free to e-mail me or leave a comment.
  • pink78 - Monday, April 7, 2008 - link

    But can any of us have something really beterr in such a small case? The Air's not that bad, and by the way, the battery life is not that bad, too, There are situations much worse!
    And by the way.. http://www.maconair.com/macbook_air_replace_batter...">http://www.maconair.com/macbook_air_replace_batter...
    You can actually replace the battery yourself..
  • Phrozt - Monday, February 11, 2008 - link

    Quite honestly, those are pretty weak browsing tests. I rarely sit on one page for any extended period of time. Usually I have quite a few tabs open at any given time (9 right now).. plus I open and close more all the time; especially when I read wiki (always finding side articles to read after I finish w/my main inquiry).

    Also, I think that video is a fairly large part of browsing nowadays... from youtube to all the crap people load on their MySpace page, to CNN, there are videos in almost all web browsing (save for strict office spaces... but then you'd have a workstation anyway, so this would not be for that scenario).

    Then you can add in high performance pages, such as WoW sites/fansites and/or other gaming sites. Since most of the world is broadband, gone are the days where most sites focus on bandwidth conservation, and therefore have no problem providing massive amounts of high res pictures, good design which may come at the cost of extra background images, and then the ever present blinking/flashing/interactive ads.

    I do not mean to poke at the tests set up by anand; they were good and consistent browsing test. However, I feel that the average user would be doing much more system intensive browsing than that outlined in this test, and therefore could probably expect much less available time in real world scenarios.
  • mynest - Sunday, October 24, 2010 - link

    We ran a typical battery of tests(you can go to <a href="http://laptopsbattery.co.uk/blog/">noteboo... battery reviews</a> to get more infomation), including games, applications, and a couple other areas. Time constraints limited the amount of testing we could get done before the NDA lift (and it's already a few hours late), but we will be providing a follow-up article with further testing and information. That said, if there are any specific areas that you'd like to see tested, feel free to e-mail me or leave a comment.
  • yacoub - Monday, February 11, 2008 - link

    http://arstechnica.com/reviews/hardware/macbook-ai...">http://arstechnica.com/reviews/hardware/macbook-ai... - HDD
    http://arstechnica.com/reviews/hardware/macbook-ai...">http://arstechnica.com/reviews/hardware/macbook-ai... - SSD
  • jim8151 - Thursday, February 7, 2008 - link

    I didn't read all the comments, so don't know if this was mentioned, but what about the cross-country flight scenario? Wi-fi and Bluetooth could be disabled, and typical activities on the plane might be watch a movie, draft e-mails or other texts, work on presentations or spreadsheets, play chess, etc. It'd be interesting to see if turning the radios off gets you closer to the 5 hours.

    The other thing is battery conditioning: might this not help to extend the battery life?
  • Anand Lal Shimpi - Thursday, February 7, 2008 - link

    I agree on doing an even more target usage case, I'll see if I can manage to whip one together. The problem is making something repeatable, I don't want to sacrifice repeatability/reliability to hit a different workload. I don't give up easily though, if I can manage something I will.

    The battery was run down completely twice before any of these tests were run. I don't believe battery conditioning has any real impact on the results though.
  • goinginstyle - Friday, February 8, 2008 - link

    Thanks for the first "real" battery tests I have seen on this notebook. I was starting to think "wtf" is Apple up to with the 5 hour statement and the first reviews showing two hour rates. Waiting on your review before I buy one now.

Log in

Don't have an account? Sign up now