Microsoft compares browser power consumption

When comparing web browsers, most folks tend to focus on features, user interfaces, standards compliance, and how quickly pages are rendered. What about power consumption, though? Microsoft’s IEBlog explores that question in surprising detail—and with fancy graphs. Oh my.

Since the testing was conducted by Microsoft, the results should be taken with a grain of salt. Not only did Redmond choose the hardware on which the browsers were run, but it also picked the scenarios under which they were tested. Those scenarios covered rendering an about:blank page to probe interface power consumption, displaying an unnamed "news site" that uses HTML4, and running a couple HTML5 demos from Microsoft’s IETestDrive site. In addition to monitoring power consumption at the system level, Microsoft measured the power draw of individual components like the CPU, memory, and graphics processor.

As one might expect, Internet Explorer 9 has the lowest overall power consumption of the bunch, which includes the latest versions of Chrome, Firefox, Opera, and Safari. The differences in power consumption are small enough to be of little concern to desktop users trying to save a few bucks on their monthly electric bills. However, notebook users striving for the best possible battery life may want to take note of the results. When rendering the news site, for example, Microsoft claims that IE9 would deliver 4:46 hours of battery life, while Chrome 10 would only get you 4:07.

Comments closed
    • HisDivineOrder
    • 9 years ago

    You know your browser sucks when you start pulling out the, “Well, well… my browser won’t use as much POWER as yours!” Haha, Microsoft. You never cease to amaze me. You complain to the EU that Google’s hurting YOUR ability to compete in search and you pull out power usage as your benchmark of why IE9 is better.

    What’s next? Going to complain to the government that OpenOffice and Linux are just too free to compete against? Publish a report proving that PS3 uses more power than the Xbox 360 and costs its users an extra $2 a year? Are things getting THAT bad that you can’t keep up? We know Windows 7 Phone didn’t work out, but hey… you guys got Kinect. You’ve got nothing new to sell with it for the last few months, but hey… you got the hardware! It only costs $75 to manufacture. Surely, that’s enough to help you stay afloat until you conscript some unsuspecting new developer to scrape together a Halo…

    • Meadows
    • 9 years ago

    I’m saddened to see they haven’t used the developer build capable of “HTML5”, instead omitting Opera from that/those test(s) entirely. Granted, it’s acceptable if you consider that the version that’s actually widely public is indeed not capable.

    The tests measure idling, and that’s bad, too. There are things like scrolling CPU usage (and framerate), or displaying/using/idling on flash-ad sites. Those are very representative of actual usage and customer satisfaction, so they should’ve received most of the attention, too. I mean, if I have a portable internet device and my browser is only idling anyway, then what do I care about battery life? I’m not even using the thing to begin with.

    • JMccovery
    • 9 years ago

    This somewhat unrelated, but, why does Microsoft insist that I install IE9 through Windows Update, when I’ve already installed it? Also, when was the last time you had to restart your computer, or every program that is running, just to install a damn browser?

    I only use IE for work-related stuff, and if I didn’t, I would uninstall it if I could.

    In this day and age, IE is the only major browser that doesn’t support WebGL! WTF Microsoft?

    Ok, I’m done.

      • d0g_p00p
      • 9 years ago

      Updating system DDL’s. If the DoJ did not hit MS with a monopoly charge, I can guess that IE would not be so integrated into the host OS. Not to turn this into a anti-Apple post but I am still amazed that MS was attacked for including a browser with their OS (ie anti-competitive monopoly) yet Apple gets ignored for including iTunes (much less Safari) and can skate away free without worry about anti-competitive charges.

      So yeah. IE = integrated into the OS so it requires a reboot after install.

        • Sahrin
        • 9 years ago

        Apple is irrelevant basically everywhere but the US with some traction for iPhone in Japan and Asia.

        • BobbinThreadbare
        • 9 years ago

        95% market share vs 5%.

        Monopolies have to play by different rules.

    • JollyPepper
    • 9 years ago

    So MS wants to paint itself green, eh ?

    IE never was a user’s browser but an ad rammer for the so called online advertisement “click” business.

    Maybe they should consider the extra amount of oxgen I use swearing when IE pulls megabyte after megabyte of useless data from the net to cram it on the display.

    • DrCR
    • 9 years ago

    If Microsoft benched it, then it must be true.

    • d0g_p00p
    • 9 years ago

    This is a joke right?

    • OneArmedScissor
    • 9 years ago

    It would be funny if old browsers had better battery life, which they surely must have tested, but neglected to mention. It reminds me of how they talked up 7’s improvement over Vista, when it turned out it still didn’t match XP *face palm.*

      • Sargent Duck
      • 9 years ago

      eh? They talked up Windows 7 improvement over Vista and they delivered. I think most would agree that Windows 7 is by far the best OS Microsoft has ever released. Unless you’re talking performance compared to XP. But then I remember everybody talking about what a memory hog XP was to 2000. Then again, Windows 95 was a memory hog to MSDOS. Thus, from taking a broad and generic statement that you made, adding my own little rant and a small dash of nerd rage, can conclude that Windows 7 is crap when compared to MSDOS’s memory usage. *face palm*. Unless you were talking about something, then I’ve gone off tangent but I’ll keep this post because I rather fancy it.

        • OneArmedScissor
        • 9 years ago

        I’m not talking about an OS e-peen contest, anyone’s opinions, or what defines a “memory hog,” just one fact that wasn’t represented entirely truthfully.

        Vista was a pretty sizable step back in battery life from XP. MS talked up the improvements in 7 as if they were “correcting” that issue, but they seemed to have decided that Vista was the new bar to measure by, not actually XP, and only went as far as to beat it by a little. It was a bit misleading, much like this test may be without knowing more about their frame of reference.

      • jstern
      • 9 years ago

      Newer operating systems bring more features than their previous versions, so you can expected them to require a system with better specs. So that’s kind of a bad example to use.

      I remember when XP 1st came out, I was sort of new to computers, and I saw that it required a minimum of 66Mhz CPU or something like that, and I thought that would greatly affect performance compared to Windows 98. The logic was that if I have a 300Mhz computer, then it’s going to steal those MHz, the same with the logic of some people when Vista came out, that since the recommended specs was like 1Ghz, then it was going to make a 2Ghz compute = to 1Ghz.

    • stdRaichu
    • 9 years ago

    You’d think that someone, somewhere at microsoft would know how to use bar graph axes to represent statistics truthfully.

    Coming next in a TR review: all FPS charts to start at 60 on the X-axis.

      • swampfox
      • 9 years ago

      The graphs are accurate, since they’re showing the *additional* power usage compared with idle. So, while the total consumption isn’t dramatically different (say, between 11w and 12w), the difference between the browser power usage seems fairly represented to me. Or did you mean something else?

      • NeelyCam
      • 9 years ago

      Marketing != truthful

      • OneArmedScissor
      • 9 years ago

      When are those bar graphs ever “truthful?”

      I’ll bite and even use your FPS chart example. They very often show every single game averaging a high enough FPS to never lag, and yet, one card will show up as “twice as fast” as another.

      There’s always a slant.

    • Silus
    • 9 years ago

    As you pointed out Geoff, this should be taken with a grain of salt, but I’m pretty impressed with IE9. It’s what I picked in the browser poll a few of weeks ago, because it is the one that showed more improvements over the previous versions and seems to have done them very well (on some instances better than the competition).
    Without further proof, it’s hard to be sure, but I don’t necessarily doubt these claims. IE9 does seem even more lightweight than Chrome, in terms of resources and CPU utilization (not on all cases of course, but overall), so it theoretically should be less “heavy” in power consumption.

    • xeridea
    • 9 years ago

    The last MS browser comparison with IE8 vs others, MS did nothing but lie, lie, lie, about everything. They knew IE8 sucked and they couldn’t post anything without lying and twisting everything around. They try to say they have this awesome search engine, but all all it does is copy Google, in search results and many of the features. They are even undeniably caught, lie about it, and copy something else.
    I can guarantee they made the tests to give them the best possible results while the others would get the worst results. What kind of testing setup uses “new site”? It could be a “new site” that they cooked up, like when they tried to cheat on the Acid tests. Look at the graph, it only goes from 10-15 in an attempt to make the differences they “found” look bigger. Reminds me of people who tell their parents they are going to “a friends” house in an attempt to not have them find out what friend, or what for.
    Oh and for features, they say they support like 99% of HTML5 and whatnot, when really they support like 70-80%, and they pass 99% of their internal development tests. I don’t trust anything coming from MS comparing their products to others.

      • bthylafh
      • 9 years ago

      Nerd rage is the funniest rage.

      • Sargent Duck
      • 9 years ago

      What Microsoft did is no different than what Nvidia has done, or AMD or Intel, or, and here’s a big one, Apple. It’s called marketing. EVERY company is going to cherry pick their tests to make themselves look better. There isn’t a single company out there that doesn’t “enhance” their own products a little when running a comparison test. It’s fine that you don’t trust anything coming from MS when comparing their products to others, just make sure you apply that same brush to every other company out there.

        • internetsandman
        • 9 years ago

        There’s looking for ways to show your product in the best possible light, like AMD and Intel do, and then there’s lying, rebranding, competitor crippling and all of the other foul marketing tactics that Nvidia and MS are doing. I’d much sooner believe (with a grain of salt of course) a benchmarking from AMD than I would from Nvidia

          • insulin_junkie72
          • 9 years ago

          [quote<]and then there's... rebranding,.. that Nvidia and MS are doing[/quote<] AMD is on the rebranding train, too, so you best scratch that one of your list.

            • NeelyCam
            • 9 years ago

            At least Intel doesn’t rebrand.

            • bigfootape
            • 9 years ago

            Nope. They don’t sell Centrino processors either, or the V Pro platform, etc, etc.

            • NeelyCam
            • 9 years ago

            Did you just miss the point?

    • leigh crusio
    • 9 years ago

    The most CPU intensive part of “web browsing” is the flash player and as far as i know its pretty much the same across all browsers.

    Static ads are fine, I can live with them without issue. For everything else there is flash block and an extra hour of browsing on my laptop.

      • OneArmedScissor
      • 9 years ago

      There is probably just as much of a difference with how Flash is handled between each browser as there is with anything else between them. Browsers have to support specific features of Flash, run their own plugin, or in Chrome’s case, it’s integrated into the browser itself. It’s not important what aspect is “most CPU intensive.” Flash itself isn’t even CPU dependent anymore.

      There’s so much that can be bounced back and forth between the CPU and GPU, so many ways to do it, and it’s only going to become more complex over time.

        • bcronce
        • 9 years ago

        Yet even Flash 10.2 can hose one of the cores on my i7 and 6950.

        Vector processing/etc can be offloaded to the GPU, but many ads have horrible code that has nothing to do with rendering.

        • UberGerbil
        • 9 years ago

        But since Flash is frequently used for ads, and specifically for ads that are rotated over time (and aren’t static while they are displayed), it’s quite likely that on many otherwise static web pages it’s the only thing that is consistently keeping much of the hardware from being idle — not just the CPU+GPU, but the NIC and everything in between. While it’s true that there are other ways to serve ads, and there are plenty of AJAX pages that are pulling data in other ways (and for other reasons) I still see a lot of web pages that in the browser amount to static HTML + Flash ads.

          • hapyman
          • 9 years ago

          Flashblock all by default… That is the way to go on any device imo.

            • UberGerbil
            • 9 years ago

            Even easier: I just don’t install it.

            • BobbinThreadbare
            • 9 years ago

            Well I sometimes want to consume video content on the internet, so your solution doesn’t really work for me.

      • Meadows
      • 9 years ago

      Flash is less CPU-intensive if you’ve updated it and use a good browser. Opera has amazingly low CPU usage even with Flash, and Chrome suspends unused Flash tabs with next to zero CPU usage, from the moment you switch away to another tab. Older browsers don’t do these things well, if at all.

    • OneArmedScissor
    • 9 years ago

    I find this really funny because they didn’t seem to make any attempt to show IE9 having an edge over Firefox, far and away their biggest competitor.

    Instead, it just makes Safari look bad compared to every other browser, which I’m sure MS are just as happy with haha.

      • NeelyCam
      • 9 years ago

      Yes.

      The results and the tables are a mess, too. The numbers in the last table don’t match anything else on the test.

      Actually, the last table is very interesting. If we assume that a regular user spends less than 50% of his/her time running MS benchmarks, all the other browsers mop the floor with IE9.

      Also, why would they weigh staring at empty pages equal to the “News Site”?

    • Sargent Duck
    • 9 years ago

    Interesting. *If* this test is true, that’s a pretty big difference, one I never would have guessed. I wouldn’t worry too much about the hardware issue, that seems like a moot point. It’s not like a particular CPU or ram is optimized to run IE9.

    The software, yeah that can be cherry picked. What I take from all of this is that TR needs to do their own independent review! Time to bust out the power meter!

      • Goty
      • 9 years ago

      It’s really not a big difference in most cases. For FF4, for example, the total difference amounts to ten minutes of battery life.

      • swampfox
      • 9 years ago

      Agreed. I think this is really interesting, and would love to see some independent reviews (particularly of mobile browsers for phones, where it might matter more). I’m particularly curious to see how legacy browsers faired (maybe the last few editions of Safari, FF3 and FF3.5, an earlier edition of Chrome – doesn’t matter too much which one since they change so often – and IE6).

    • dalisam
    • 9 years ago

    It’s a shame I can’t use IE9 on my XP netbook where this would be more helpful

Pin It on Pinterest

Share This