The other hidden cost of running Windows on a Mac: battery life

The other hidden cost of running Windows on a Mac: battery life

Summary: Thinking of buying a MacBook so you can run Windows on it? You might want to think again. A year's worth of data suggests that you'll pay a steep price in battery life.


There’s no denying that Apple makes handsome hardware. Its MacBook Air line, for example, started out as an underpowered and overpriced novelty, but thanks to SSDs and faster Intel CPUs it’s now a powerful little machine.

Apple’s hardware designs are so handsome, in fact, that some people are willing to buy a Mac just to run Windows. But is that a good idea? Maybe not—at least not if you want to get the power and battery life you expect from the components inside.

The problem is Apple’s Boot Camp, which is the only supported way to run Windows directly on Apple hardware. The Apple-supplied drivers are substandard, and they don’t allow Windows to take full advantage of the underlying hardware. I’ve written about this problem previously in terms of its impact on disk and graphics subsystem performance.

And now there’s a body of evidence that documents another hidden cost of running Windows on a Mac. You’ll lose a substantial amount of battery life by doing so.

The good folks at Engadget have been using a consistent test methodology on MacBooks and Windows-powered Ultrabooks for the past year. That collection of data makes it possible to compare machines easily, as they’ve done in a recent review of the new Lenovo IdeaPad U310.

What’s especially interesting is that their battery-life tests on the brand-new 2012 MacBook Air and its 2011 counterpart were done with OS X and with Windows 7. (Engadget’s 2012 MacBook Air review confirms that the Windows tests were performed using Boot Camp.)

I’ve taken the liberty of borrowing that data and refactoring it into a table that ranks the current crop of Ultrabooks and MacBook Airs according to battery life. The results are eye-opening.


The first conclusion to draw from that table is that the current crop of Windows 7-powered Ultrabooks do a very good job in terms of battery life. For now, at least, Apple is the one playing catch-up.

As you can see, the 2012 MacBook  Air running OS X gets very good battery life, although three Windows-powered Ultrabooks (including two Samsungs) outlast it by 7-14%.

The 2011 MacBook Air running OS X is literally in the middle of the pack, with eight current Windows-powered Ultrabooks able to outlast it. The best performer, the 2012 13-inch Samsung Series 9, can run for 90 minutes longer on Windows 7 than a year-old MacBook Air running OS X.

But the really interesting figure for me is how poorly those MacBook Airs do when running Windows 7 under Boot Camp.

The 2012 MacBook Air, which is near the top of the heap running OS X, delivers pitiful battery life on Windows 7. The difference is more than two hours of productivity: 6:34 on OS X, 4:28 on Windows.

What’s really fascinating, though, is to compare performance with similarly configured Ultrabook designs from multiple PC OEMs. A year ago, those comparisons were difficult. Today, the wide availability of Ultrabooks makes it possible to see just how weak Apple’s Boot Camp drivers really are.

Nine Ultrabooks from six different manufacturers, running Windows 7 out of the box, deliver battery life that is at least one hour greater than the brand-new 2012 MacBook Air running Windows 7 under Boot Camp. The new Samsung Series 9 runs more than 2-1/2 hours longer than that same MacBook Air running Windows 7.

I don’t blame Apple for this terrible performance. They’ve focused their engineering resources on their own hardware and their own operating system. For Apple, Boot Camp is a tool to use occasionally, when you need to run a Windows program without virtualization software getting in the way.

If you're a Mac user who occasionally dips into Windows, this isn't an important factor for you. 

But these numbers should serve as a warning to anyone who plans to use a Mac with Windows as a full-time platform. The hidden costs are much higher than you think.

Topics: Hardware, Apple, Hewlett-Packard, Intel, Lenovo, Mobility, Samsung, Windows

Kick off your day with ZDNet's daily email newsletter. It's the freshest tech news and opinion, served hot. Get it.


Log in or register to join the discussion
  • The other issue:


    Apple writes crappy drivers because of one thing: they don't use the industry standard 64-bit UEFI 2.0 for their EFI. Instead, they use an unoptimized BIOS software emulation for non-OS X operating systems, and as such, the drivers don't take advantage of a lot of PC standards like AHCI or full-speed PCI-e in anything except their own OS. The PnPID's that they use don't match generic hardware using the same vendor's chips so you need to get almost all drivers direct from Apple and they don't update them enough to make it worthwhile running Windows on their machines.

    If you want a decent machine, there is far better support from hardware vendors that preinstall Windows.
    • i am a bit confused

      If i understood it uses industry standard uefi 2.0 but the drivers emulate a unoptimized bios?
      I have to admit I don´t know whether its hard to emulate a full bios WITH AHCI or full-speed PCI-e, but you have to consider that windows first supported UEFI in windows 7 and maybe somebody needs vista/xp etc. Therefore apple has to emulate a Bios and can´t let the OS "see" the UEFI.
      • Not for older OSs

        Boot Camp only supports Win 7 on current hardware so that's not the reason. If you were cynical you'd think it was so OSX looks better, particularly when lazy journalists use Macs to show how much slower windows is than OSX.
        My boss insists on buying expensive iMacs for us to run windows on (massive fanboy) but the 2.4GHz Core 2 duo is actually SLOWER than my old 2.1GHz Pentium M (the predecessor to the Core) Dell laptop!
        • But, PCMag says. . .

          But PC Magazine has been touting Macs as some of the best machines that run Windows for years now and some years have said the Mac was the best, the fastest. Maybe you could show those fanboys the errors of their ways. ;-)
      • It's all in the bootup

        Apple doesn't do proper 64-bit UEFI 2.0 booting.

        Nobody has been able to do UEFI boot from Vista SP1 or Windows 7 media. When they do the BIOS emulation, a lot of the functionality is disabled (AHCI for the hard drives, so they don't support NCQ, etc.), video cards don't run at full PCI-e speeds, etc. The drivers and hardware support for Windows on Mac is just poorly done, and when you need vendor support for something that is provided half-assed, especially in a business environment, it's wasted time, and time is money.
      • A couple of points you're wrong on

        Apple doesn't support real UEFI 2.0, otherwise Windows 7 would boot off the media. They are still using proprietary EFI.

        And the first time Microsoft supported UEFI was in Vista SP1, not Windows 7.

        The BIOS emulation isn't done properly on Mac's. If they did, we wouldn't have these problems. If they supported native UEFI 2.0, we wouldn't have these problems with Vista SP1 and Windows 7 64-bit either.
    • In reality

      OS X (a UNIX-based operating system) is much more efficient, and less power-hungry than Windows. This is why OS X runs so much longer than Windows on the same hardware.

      Ed Bott compares other laptops to the MacBook Air, with total disregard for the different specifications of the devices.

      Laptops with (thicker and heavier) larger batteries, slower processors, or lower resolution displays, will obviously have longer battery life than the MBA. This is why the 13" MacBook Pro, with a larger battery, has a MUCH longer battery life than the 13" MacBook Air.

      It's very easy to give the negative results you want to show, by simply comparing Apples to oranges. ;-)
      Harvey Lubin
      • Same ol' same ol'

        Harv, it's Ed Bott. You know, the same guy that used to call his column the Microsoft Report or something like that. The name has changed. He hasn't.
      • Wow. Skewed logic. So bad it hurts to read it.

        First off, here is what you say:

        OS X (a UNIX-based operating system) is much more efficient, and less power-hungry than Windows. This is why OS X runs so much longer than Windows on the same hardware

        Here is what the facts say about the tests:

        2012 MacBook Air running OS X gets very good battery life, although three Windows-powered Ultrabooks (including two Samsungs) outlast it by 7-14%

        Your response is to say:

        Laptops with (thicker and heavier) larger batteries, slower processors, or lower resolution displays, will obviously have longer battery life than the MBA.

        But how much heavier??? How much thicker??
        The Macbook Air in its lightest form weighs 2.38 lbs., in its heavier form, 2.96 lbs
        The Samsung series 9 13”weighs 2.5 lbs.
        The Macbook Air is 1.7cm thick
        The Samsung series 9 13” is 1.27cm thick

        From what I see the weight difference is not significant, in the case of the thickness, the Samsung wins, and in the case of the more powerful Macbook Air it even weighs more then the Samsung 13”.

        Of course if you start comparing the Macbook to the Samsung 15” we are into a whole new ballgame, as you say comparing Apples to oranges when it’s a 13” screen to a 15” screen. I hope you have no doubt that for anyone who finds with screen size, bigger is better, it would be very unfair to compare the lowly 13” Macbook to a 15” Samsung. So not only do they get that bigger screen, they are getting that additional huge amount of battery time. So except for only a few people, who for some reason really cannot take an extra pound or so of weight, the extra weight is hardly worth mentioning if they are getting a noticeably bigger screen they want and even more battery time.

        The bottom line for any hardware comparison you are dealing with whatever entire packages you are comparing. Screen size, weight, performance figures and user experience. In a comparison of this kind its plain foolish to pretend that for most people once they decide on the form factor, in this case ultra slim lightweight notebooks, the biggest selling point of all will be the particular operating system. If you’re an OSX fan its going to be horribly difficult to convince you that the 15” Samsung is an over all better choice for any reason because its simply not an Apple product running OSX. And visa versa, if you want a Windows product its going to be just as brutally difficult trying to convince someone that even the 13” Samsung, almost as light but thinner and with better battery life than the Macbook is not a better choice than the Macbook.

        So the whole point of this article was not to trash Apple at all, but to point out that if what you want to do is to run Windows, its plain foolish to buy a Macbook Air to install Windows on it because your not going to get a significantly lighter laptop than the Samsung for example, your going to get a slightly thicker laptop if you go with the Macbook, and your certainly going to get far worse battery life on the Macbook if what your intending to do it run Windows on it.

        Your ranting dosnt seem to change those facts in any rational way.
    • For example:

      The Dell XPS 13 has about a 10% battery life improvement over the 13" MacBook Air running Windows, but then again...

      The Dell XPS 13 has a slower i5 processor than the 13" MacBook Air
      The Dell XPS 13 has a slower integrated GPU than the 13" MacBook Air
      The Dell XPS 13 has a lower resolution display than the 13" MacBook Air

      It is these lower spec differences that give the Dell XPS 13 the slightly longer battery life... it has nothing to do with any other imaginary voodoo.
      Harvey Lubin
      • Yeah right...

        "The Dell XPS 13 has a slower i5 processor than the 13" MacBook Air
        The Dell XPS 13 has a slower integrated GPU than the 13" MacBook Air"

        So what you are telling us here is how crappy OSX is!! :-)

        All modern OS (OSx and Win7) cuts down the processing speed on the CPU & GPU to the bare minimum. At Least Windows 7 does this. So If a Mac using Core i7 needs more juice to run then its only because OSx is stupid enough to not throttle it down. If it did that the power consumption would be barely noticeable between i5 and i7. So thanks for pointing that out. Another reason to buy Windows laptops. :-)
        • No

          This means that the only way to prolong battery life is to use less powerful hardware.
          If Ed Bott compare MacBooks with PCs with less powerful hardware then of course the PCs will have longer battery life.
      • By that logic...

        The 2011 MacBook Air should have gotten much better battery life than the 2012 MBA, which has more powerful hardware. But wait, that's not true!

        Oh dear, maybe it really is true that the different architecture of the Mac hardware, coupled with mediocre drivers, causes Windows to run poorly compared to similarly spec'd PC hardware.
        Ed Bott
  • very interesting! what about running windows in a VM?

    This is very useful to know. I am thinking about buying a Mac but I would need to run Windows. would you have useful stats concerning the battery life when starting OSX and doing nothing but running windows in a VM? Is the battery life as poor as running windows under boot camp or is it the same as we would get by running OSX?
    • Virtualization is a resource-intensive task

      Logically, one would assume that battery life in a virtualized environment would be somewhat less than running only native OS X apps in OS X. Just how much less, though, is anyone's guess. I would love to see Engadget address this topic directly.
      Ed Bott
      • Well with 64bit multi-core i7s

        Hyper-V support, 8 to 16GB+ of RAM and specs that would make the 32bit machines of yore blush with envy, that shouldn't be much of an issue anymore.

        Get with the program, Ed
      • If battery life is deficient

        It's because not enough is being invested in new battery technology, the Achilles heel of mobile technology.
        • There's two sides to battery life

          Battery life is determined by both how much power the battery can hold and how slowly the system drains that power. And the point of the data is that while both Windows and Apple have done a lot of work making their OS's function with minimal power usage, part of that efficiency is based on taking advantage of what modern hardware has been made to be able to do, and the data makes it clear that the quality of Apple's Windows drivers greatly impair it's ability to ramp up the os's efficiency to it's maximum.
          • Did you have a point to make?

            Any VM you run on battery power (regardless of the OS) is going or eat into it, big time.

            That still doesn't negate what I said about current battery technology.

            Did you STILL have a point to make?
    • Do you really need Windows

      It has a big impact on your battery live AND your performance (especially on windows)....i don`t know exactly how much, but its enough to notice it. You have to consider that you are running 2 OS + the emulation software ALL the time.

      If there is a alternative program out there, use it!

      I know enough persons who emulate windows just because they like outlook on windows more, which is really stupid i think.