17
Mar
mike-tyson
Last Updated: March 18th, 2011

Update: After getting a lot of initial press, experts are weighing in and crying foul with the methodology of this study. Apparently the custom app the researchers used relied on Apple's UIWebView, which doesn't utilize recent optimizations made to the actual Safari browser.

John Gruber of Daring Fireball says:

"That's not to say it isn't interesting that Android's WebView for apps is faster than iOS's UIWebView for apps, but it just isn't true that these results are indicative of anything regarding Mobile Safari's performance. It's easy to see that Mobile Safari is faster than UIWebView - just run something like the SunSpider benchmark twice, once in Mobile Safari and once in any app from the App Store with a web content view. On my iPhone 4, Mobile Safari runs SunSpider almost three times as fast as an app using UIWebView."

We would now consider this study to be flawed at best, if not completely meaningless. We would, however, look forward to a similar test done under conditions that could lead to an accurate measurement.

Web analysis site Blaze has conducted an extensive study to compare the browser loading times of Android and iPhone, and we think most of our readers will be very pleased with the results. In a dominant showing, the Android browser was determined to be, on average, 52% faster than the iPhone's.

Blaze loaded over 45,000 web pages from 1000 web sites (real sites, mostly those of Fortune 1000 companies), using an iPhone 4 running iOS 4.3 and a Nexus S running Android 2.3 'Gingerbread.' Each page was loaded multiple times on different days. Their study used a custom benchmarking app that measured page load time. Stock browsers were used for each OS. While the average speed of the Android phone was 52% faster than the iPhone's, the Nexus S also handily beat Apple's flagship device 84% of the time in head-to-head browser loading races.

chart_fastestmedloadtime

chart_winlossloadtime_v2

When broken down into a comparison of mobile sites vs. desktop versions, Android's primary advantage became apparent. Mobile versions of sites loaded in roughly the same time for each phone (two seconds on average) but for regular sites, the Nexus S loaded them faster by over a two second margin (2.1 seconds to be exact, compared to 4.3 for the iPhone).

The study also found that faster JavaScript engines, touted in both iOS' and Android's latest iterations, didn't add much improvement in browser speed. Blaze's researchers hypothesized that the "improvements" were largely intended for benchmark purposes and don't actually add any advantage in real website browsing conditions. iOS 4.3 showed no improvement over the earlier 4.2 and Gingerbread showed only a marginal boost over Froyo.

While these results likely comes as little surprise to many Android users, many Apple fans and other less Android-centric smartphone customers may be surprised to hear that Apple's software and hardware do not, in fact, "just work" better in this case. We can only wonder how Steve Jobs will find a way to frame this one as an advantage for Apple when it comes time for his next grand unveiling.

You can head over to the source link for more details about their findings.

Source: Blaze via Gizmodo and Daring Fireball via Network World

Will Shanklin
Will's typical, run-of-the-mill story is that of 'classically-trained actor turned Android smartphone and tablet writer.' If you catch him quoting Shakespeare, it's not because he misses it, but because he desperately wants his Masters in drama to count towards something.

Sir William dwelleth in the fair haven
Chicago; with his fair maid'n Jess and his
trusty cur, Ziggy.
  • http://www.AndroidPolice.com Artem Russakovskii

    Kind of expected, although results are heavily dependent on the hardware used, so it's not just "Android Browser" but rather "Android Browser on Nexus S." I'm sure Droid's Android browser is not nearly as fast and Droid Bionic's could be even faster.

  • http://mattters.com Mike

    It is a great personal pleasure to let you know that the several million visitors to Mattters over the last year have selected your blog as one of their favorite blogs of 2010 (7th on the Android channel!).

    You can see your blog, along with the other contenders, on the Android channel at http://android.mattters.com.

    P.S. We made a cute little '2010 Blogs that Mattter' award that you can pin up on your website if you like those sort of things (you can see it at http://android.mattters.com/awards).

    Oh, and if for some reason or another you ever think your blog should be on a different channel, or even multiple channels, or perhaps none at all, (or you just want to chat!) just send me a quick email at michael@mattters.com.

    All the best,
    -Mike
    http://mattters.com
    Mattters - Follow Your Interests!

  • James

    What? Are you stupid? It still "just works" dummy, it's just not as fast. How do you get those two things mixed up?

    • http://schpydurx.livejournal.com ProfessorTom

      He's a fanboy. By definition, fanboys don't think.

      • James

        Oh. Got it. Thanks.

        • Eric

          Are you a second ProfessorTom account? Or is android police reporting on things that threaten iPhone users enough that they are all going to feel the need to troll.

          Some people need a life. He made a joke. You're on a site with android in the title. If I went to TiPb, or other iPhone blogs, I'd expect to see jokes about android. In fact, I have. Difference is, I don't go trolling.

  • James

    @Eric Are you talking to me or ? I think it's pure fanboyism - I didn't see any joke at all.

  • JC

    So are you guys going to update this article now that it's been debunked? Even the original tester's link you posted has been updated to admit the testing was flawed. They tested their custom app instead of using MobileSafari, which has caching enhancements, asynchronous multithreading, and a different JavaScript engine -- things not present in the custom app.

    • Coldman

      Looks like it got updated.

Quantcast