Android 2.2 (Froyo) versus iOS 4: the browser showdown (video)
By Richard Lai 
posted July 7th 2010 1:56PM
posted July 7th 2010 1:56PM
Update: for those who were concerned about the battery affecting the Nexus One's performance, we did use Android System Info to verify that the CPU was still clocked at 1GHz. We were also able to reproduce the same results with a full battery. Either way, it's still a win for Android.
Our test candidates were a 16GB iPhone 4 (with a shameless color mod) and a Nexus One rocking the official OTA 2.2 update and Flash 10.1. Naturally, we cleared out the cache files on both devices prior to each trial. Out of the five desktop-version websites that we tested for load time, three of them -- BBC News, gdgt and The Onion -- repeatedly produced a tie between the two phones, whereas the iPhone 4 consistently loaded Engadget about two to three seconds faster, and the Nexus One about one to two seconds faster with New York Times. Here's our video of one of the trials:
At this point we thought: would things load quicker with Flash uninstalled? Turns out it's a small yes -- this time the Nexus One consistently loaded Engadget, New York Times, gdgt and The Onion a tad quicker than the iPhone 4, whereas BBC News produced a tie. Still, this is far from the massive performance jump that Ars Technica's benchmarks suggest (not that we're saying the site did it wrong; these every-day sites just don't rely heavily on JavaScript), but considering Froyo running Flash still matches iOS 4's Safari loading speed, Android's just about won this two-man race. Apple, your move next.
@Glock Let me just quote for you:
"And boy, it sure looks like it was worth the wait -- Ars Technica's JavaScript benchmarks show that not only is Froyo's browser almost three times faster than its Éclair counterpart, but it also beats iOS 4's Safari by at least two-fold. That said, numbers alone don't always reflect real-life performance -- especially with Froyo supporting iPhone's much-missed Flash -- so we went ahead and conducted our own browser speed test."
Pretty sure I'm not the one who has issue with reading.
@AJerman
You pwnd him just like froyo pwns iOS4
Too bad I don't visit full sites regularly on my phone...I typically aim for the Mobile sites since they are esier to read and give me the info i'm looking for.
I don't like pinch to zoom on the iphone becuase it doesn't do text reflow like Opera Mobile
@daguila29
I guess that you have an iphone but full sites on the nexus and i assume other android phones are just as easy to read as any mobile site. The double tap feature will zoom in and realign the text fitting it so you just have to scroll up and down.
There is a big difference between a test of loading pages and a test of using pages. If there isn't a lot of JavaScript being executed on page load, the JS engine differences will be largely moot. The test would be to actually interact with a complex JS-based webapp, not just loading a page.
@meatless Yes, the whole point is that the benchmarks don't mean much to the average users.
@Richard Lai
Personally, I'd say the point is that Apple will now STFU about Safari being the best mobile browser. :D
@Richard Lai
I think this is a very poor response. You cannot run a few web pages and then make the statement that increased Javascript performance will not matter to the average user. You are making assumptions on what sites the "average" user visits, how many times they visit them,etc.
Ars is testing Javascript benchmarking, and nothing more. So therefore their claim is valid. Your test however, is not only invalid, but makes a HUGE error in judging which websites the "Average" user will visit. You are a technical blog and should take greater care in how your tests are conducted and the assumptions made during that testing.
@Richard Lai There are already a bunch of "real world" sites heavily relying on javascript. Why didn't you test those? Purposely dodging the question?
Did I say thank you, yet? No?
Thank you, Engadget.
I love you Nexus one
iphone 4 is definitely the first iphone to catch my attention... but i'm definitely going to stick with the android OS. I see a lot of potential in android. ;]
Seems to me that they are the same. He's always pressing the iphone first though very slightly.
I guess my life isn't as important as everyone elses...I'm happy as long as it loads in a decent amount of time. Also, taking an extra few seconds to load with full Flash is worth the speed tradeoff if you want to call it that.
Are you saying that those web pages are purely JavaScript? This is what you seem to be referring to when mentioning the Ars Technica tests...
Damn. I figure N1 was faster, but surprised to see it beat the iPhone even with low battery and Flash. Guess Google was right about the fastest mobile web browser.
Now image Droid X + Froyo 2.2
:O
So the nexus ran the 200 meter dash(flash enabled) and the iphone ran the 100 meter dash(flash-less) and still won??? Jeeze! I need this on my evo pronto htc/Sprint!
Engadget writers/editors: You guys ARE aware that javascript rendering performance and HTML rendering performance are two different things right? PLEASE PLEASE PLEASE get a tech lead on your payroll, preferably someone who is actually programmer, much like Nilay is a lawyer. There are way too many layman's mistakes made when you guys write up tech pieces like this.
@avinash240 We never said the two are the same -- we're saying the benchmarks don't reflect real-life performance, or that they'd have a huge impact on the average user.
@Richard Lai You guys do an excellent job of reporting. However, when it comes to things like this, there are usually holes in your logic/reporting. And that's ok, that's what Ars Tecnica and Anandtech is for.
I'll use this an article as an example, you've reported a story sparked by a Javascript benchmark test; yet you're test sites are heavy on HTML. That's like me testing the straight-line performance of two muscles cars by putting them on a road coarse with a lot of turns. There are much better examples(Facebook anyone?), and a programmer could have pointed you guys to these examples of "Real world" Javascript heavy sites, where you would see a "Real World" difference.
My Nexus runs allot slower when the battery is low
Why would you even compare with Flash on?
@ekeefe41 We also compared with Flash off!
No need to uninstall Flash. Just go into settings and make it "On Demand" so it only loads Flash content when you tell it to.
Nexus loads page WITH flash in the same time that iPhone just loads the page. Nexus is faster when it doesn't load the flash (just like iPhone never does).
How does the editor come up to 'No performance increase" conclusion? And it's CLEARLY in your own video! I don't get it...
@pretol anyone that gives apple even the SLIGHTEST bit of credit is an apple fanboy, i guess engadget want to be seen as unbias
anyway, yeah, the nexus one is clearly faster, another victory for android!
damn, a 7 month old phone out performing the brand new iphone 4? lol apple
I'd like to see this on an Evo with the Froyo ROM. It's a hack and I bet it'd still run faster than iOS 4.
Very nice. Its amazing seeing how fast and powerful phones are becoming. Only leaves it to imagination what could happen in coming months.
LOL. I like how they load engadget on the iPhone 4 and the Nexus One, and the nexus one takes like 2 seconds longer, but things are MOVING on the page vs. having blank white spaces with little icons in the middle. If I have to wait 2 more seconds to get an actual website, I can handle that.
How about the samsung galaxy S??
@jdavidsson
It is going to be slower, until it gets updated to froyo.
Android moving about in the flashest of flashes!
Is it? Yes, it's a WIN! *checked flag started waving seconds ago*
What about having the flash plug-in set to 'On Demand'? just wondering if that gives it the same as having the flash plugin not installed.
Personally I think the on demand should be the default for the flash, or at least make it easier/clearer to adjust to the average user. Just put flash in the app list with a setting button and a url link to their m.flash site or something.
2nd place is the 1st loser.
Well my initial thought was that flash was slowing down the N1 until I watched the second video which confirmed it. Not sure why there is a reference to arstechnica's benchmark which is for sort of unrelated to page loading speed.
Overall I would say for both devices as they appear on the video that users wouldn't see much of a difference in loading time.
If anyone wants to see an example of exemplary mobile browsing on a just pocketable device, fire up a Dell Streak. Even if it is running 'droid 1.6, it's a delight after the overly small screen on the iPhone.
@Oflife
That's exactly the way I feel using my EVO 4G next to my friend's iPhone 4. We were doing speed tests and quality tests for AT&T 3G vs Sprint 3G and they were dead even but he had to do a lot more panning and zooming than I did. Oh and then I turned on 4G and trounced him :)
I see the engadget/apple dickriding goes both ways.
I am amazed to hear the spin from the fanboys.
It's clear that with or without Flash, both these phones are neck-and-neck in terms of real world performance.. I don't think there was more than 1 second difference for any of the tests and most of them were less than that.
LOL what a waste of battery, time and performance with flash. I'm amazed you faux-nerds want it so bad.
@MysticMaven
lol, is the iSheep jealous of features his glorified PDA lacks?
btw, the iphone is rendering at 4 times the resolution on the iOS4. Just thought i'd point that out before everyone gets too excited..
Remember to rerun this test when Froyo hits Droid X in about a month. This way it will be two phones consumers are likely to buy. Not many are buying the aging Nexus Ones other than geeks and devs.
@rolanddd I agree. The Nexus One uses last gen technology (snapdragon) whereas the iPhone's A4 is more similar to Droid X's OMAP SoC.
I hate these test and comparisons because all of does is lead to a bunch trolling nonsense!
@PrettyBoy droid Cheering about a 1 second advantage loading a large complex page like Engadget's is about as silly as it gets...
Engadget loads quicker on an iPhone...Even Engadget's servers are Apple fanboys.
These aren't even that much faster than my iPhone 3G.
@salane89 I agree. I used to use an iPhone 3G and I have an EVO now which I think is overall a better fit for me. I'm not sure if there's something wrong with your phones or what not, but I'm performing the same tests, and my iPhone 3G and EVO both are loading faster than the iPhone4 and Nexus Froyo.