Page 1 of 2 1 2 LastLast
Results 1 to 10 of 13
Like Tree10Likes

Thread: Ranking the Android Browsers

  1. #1
    Jazz
    Member #
    20215
    Join Date
    Jan 2012
    Posts
    30
    Liked
    10 times

    Ranking the Android Browsers

    RANK
    Tier 1
    #1 Dolphin Beta
    #2 Dolphin (w/Jetpack)
    Tier 2
    #3 Chrome
    Tier 3
    #4 ICS Browser+
    #5 Firefox Beta
    #6 Firefox
    #7 Ninesky
    #8 Maxthon for Tablet
    #9 Boat Browser
    #10 Dolphin for Pad
    #11 Maxthon
    #12 Skyfire
    Tier 4
    #13 Opera Mobile
    #14 UC Browser


    BROWSER BENCHMARKS
    Google Octane Higher= better
    Google's newest benchmark comprehensively testing Javascript. This benchmark supercedes (and subsumes) the old V8 standard. Websites are continuing to show V8 mainly for legacy comparisons.
    RightWare BrowserMark Higher= better
    Probably the single most referenced benchmark for browser performance in the industry.
    Apple Sunspider Lower= better
    The link is to Google's modification of Apple's Sunspider benchmark for cross-platform benchmarking. This, Octane, and Browsermark are the most commonly charted benchmarks for overall analysis of browsers. This trio forms the standard.
    Mozilla Kraken Lower= better
    About 2 years old now, but it's massive, and offers Mozilla's benchmark against which they most acutely test and develop their own software: important, IMO, as Firefox is the most significant open-source competitor to Chrome in the desktop market.
    Futuremark Peacekeeper Higher= better
    A recent benchmark gaining attention with a real focus on graphics performance in browsers. The first number is the overall score. The second number, in parentheses, is its highly simplified score for a browser's HTML5 capabilities. I included this benchmark because the results varied considerably from the otherwise mostly consensual benchmarks above.
    HTML5Test Higher= better
    A benchmark that focuses strictly on the HTML5 capability of a browser as this is the upcoming standard for the internet replacing Adobe Flash and pretty much every other form of virtual media.
    Facebook Ringmark (R.2 > R.1 > R.0)
    Facebook's new benchmark. It's different than the above benchmarks in that it doesn't offer a numerical score, but instead adopts a "pass/fail" approach to whether or not a browser can handle all the new types of coding that highly sophisticated websites like Facebook use. It tests the most critical scripts first in the innermost ring and moves outward. Ergo, any amount of failures in Ring 1 is a stronger performance than a performance that doesn't get past Ring 0.


    ICS BROWSER
    Octane: 995 {Incomplete}
    Browsermark: 99181
    Sunspider: 2263
    Kraken: 50,565
    Peacekeeper: 317 (1/6)
    HTML5Test: 287
    Ringmark: R.1 (35 failed, 87 passed)

    CHROME 18
    Octane: 1152
    Browsermark: 95,891
    Sunspider: 1904
    Kraken: 29,024
    Peacekeeper: 354 (1/7)
    HTML5Test: 390
    Ringmark: R.1 (9 failed, 134 passed)

    OPERA MOBILE 12.1
    Octane: 450 {Incomplete}
    Browsermark: 58,601
    Sunspider: 1747
    Kraken: 76,439
    Peacekeeper: 281 (2/7)
    HTML5Test: 406
    Ringmark: R.0 (8 Failed, 93 Passed)

    DOLPHIN 9 (w/Jetpack)
    Octane: 1262
    Browsermark: 99,539
    Sunspider: 1592
    Kraken: 26,079
    Peacekeeper: 466 (2/7)
    HTML5Test: 492
    Ringmark: R.2 (48 Failed, 182 Passed)

    DOLPHIN BETA 1.1
    Octane: 1319
    Browsermark: 105,815
    Sunspider: 1589
    Kraken: 26,087
    Peacekeeper: 467 (2/7)
    HTML5Test: 482
    Ringmark: R.2 (47 failed, 184 passed)

    DOLPHIN FOR PAD
    Octane: 1004 {Incomplete}
    Browsermark: 94,952
    Sunspider: 2224
    Kraken: 48,468
    Peacekeeper: 323 (1/7)
    HTML5Test: 287
    Ringmark: R.1 (35 failed, 87 passed)

    FIREFOX 16
    Octane: N/A
    Browsermark: 59,440
    Sunspider: 1838
    Kraken: 36,566
    Peacekeeper: 263 (4/7)
    HTML5Test: 388
    Ringmark: R.1 (11 failed, 136 passed)

    FIREFOX BETA
    Octane: N/A
    Browsermark: 56,677
    Sunspider: 1892
    Kraken: 36,945
    Peacekeeper: 265 (4/7)
    HTML5Test: 398
    Ringmark: R.1 (11 failed, 136 passed)

    SKYFIRE 4.1
    Octane: 1014 {Incomplete}
    Browsermark: N/A
    Sunspider: 2316
    Kraken: 49,728
    Peacekeeper: 326 (2/7)
    HTML5Test: 269
    Ringmark: R.1 (36 failed, 86 passed)

    MAXTHON 2.7
    Octane: 1007 {Incomplete}
    Browsermark: 71,619
    Sunspider: 2231
    Kraken: 50,096
    Peacekeeper: 325 (1/7)
    HTML5Test: 287
    Ringmark: R.1 (35 failed, 87 passed)

    MAXTHON FOR TABLET 1.7
    Octane: 1023 {Incomplete}
    Browsermark: 90,352
    Sunspider: 2340
    Kraken: 49,839
    Peacekeeper: 335 (2/7)
    HTML5Test: 287
    Ringmark: R.1 (35 failed, 87 passed)

    UC BROWSER 8.4
    Octane: 401 (Incomplete)
    Browsermark: 41,619
    Sunspider: 2528
    Kraken: N/A
    Peacekeeper: 173 (1/7)
    HTML5Test: 294
    Ringmark: R.1 (47 failed, 96 passed)

    BOAT BROWSER 4.6
    Octane: 1037 {Incomplete}
    Browsermark: 90,881
    Sunspider: 2244
    Kraken: 48,852
    Peacekeeper: 322 (1/7)
    HTML5Test: 287
    Ringmark: R.1 (36 failed, 86 passed)

    NINESKY 2.5
    Octane: 1018 {Incomplete}
    Browsermark: 92986
    Sunspider: 2295
    Kraken: 48,746
    Peacekeeper: 328 (1/7)
    HTML5Test: 287
    Ringmark: R.1 (38 failed, 84 passed)
    zAlbee, janner43 and lidow like this.



  2. #2
    Jazz
    Member #
    20215
    Join Date
    Jan 2012
    Posts
    30
    Liked
    10 times
    I ran these tests on my Asus Transformer TF101 running on Android 4.0.3 (Ice Cream Sandwich). The reason is that all the articles in the industry compare performances by pitting device against device. For instance, they're most interested in comparing the iPhone vs. Galaxy S3. They don't really ever analyze the performance of these different browsers on a single device. With all the hullabaloo Google has caused by removing Flash support from the Play market for devices on Android 4.0 or later (finally joining Apple in forcing users to adapt to the eventual extinction of Adobe Flash in favor of HTML5), I thought now would be the most salient time for users to consider adapting to a new browser for the upcoming standard. So I ran these benchmarks. I ran each of the speed benchmarks at least three times and took the mean.

    I've long been a Dolphin user because of gestures, but I'd noticed in the past couple of months on these forums and at XDA everybody raving about how fast and slick Dolphin had become since the Dolphin Jetpack update was released on the Play market. Well, these benchmarks justify the hype. It won every single benchmark by a massive margin. Dolphin absolutely blew everything else away. It even destroyed Chrome at Google's own Octane benchmark, and was the only browser besides Chrome to complete all tests on that benchmark. It was the only browser I tested that reached Ringmark's outermost ring.

    Keep in mind these tests mainly benchmark just HTML5 capability and speed. Flash is a separate issue. Also, I'm aware that apart from speed and Javascript capability, most users choose their browser with greater consideration for layout design and specific features. Some browsers let you switch between tabs using the volume buttons; some let you shake the phone/tablet to switch tabs; some have tab menus on top and some have a side menu you can pull out; some automatically remember your last tabs visited when you re-open the program...those kinds of things. Nevertheless, regardless of the fact that I found Opera Mobile's layout very appealing, I couldn't stand how lumbering it was on my Transformer. It's terrible scores reflected my experience. It's hard for me to fathom how it's the most popular mobile browser in the world.

    I didn't run benchmarks for the "Mini" browsers because those are very limited in regards to rendering capability. They're built purely for speed, so judge them on that; however, I should mention that although it offers a "mini" version, the UC Browser itself felt to me like a glorified mini browser. When you first open the application it asks if you'd like the app to compress server data before sending it to you to make browsing more speedy. Perhaps, then, it's the most capable of all the mini-type browsers for those seeking maximum speed without running into graphically deformed pages from a lack of capability. The major mini Browsers are Opera Mini, Dolphin Mini, Boat Browser Mini, and UC Browser Mini.

    Also know that you can visit the above benchmarks to test your desktop browser. Right now Chrome is the dominant desktop browser in these benchmarks; however, I personally still use Firefox for some specific add-ons that I find critical to use. "So it goes" with preference and utility.
    Last edited by Madmick; 11-14-2012 at 02:41 AM.
    Jeffrey, Mimo, janner43 and 2 others like this.

  3. #3
    Administrator
    Supporting Member

    Member #
    2531
    Join Date
    Jun 2011
    Location
    Thailand
    Tablet
    TF101
    Posts
    8,610
    Liked
    1000 times
    Nice job... Thanks for sharing..

  4. #4
    Jazz
    Member #
    20215
    Join Date
    Jan 2012
    Posts
    30
    Liked
    10 times
    *Reserved*

  5. #5
    Starscream
    Member #
    27085
    Join Date
    Feb 2012
    Tablet
    TF101
    Posts
    327
    Liked
    75 times
    Very interesting. Few comments:

    1. Graphs might be easier to grok than lots of numbers.

    2. Please also give standard deviation with your mean. A difference of 10 pts is meaningless if the Std Dev is 25 for example.

    3. Thinking aloud... I believe many of these browsers share the same rendering engine as stock, and only adding a custom skin (Boat, ICS+ for sure). You can tell which ones by looking at the capability scores for HTML5 Test (287) and Ringmark (35 fails) which are identical (or nearly). Based on this, looks like ICS+, Dolphin for Pad, Maxthon/Maxthon for tablet, Boat (1 more fail), Skyfire (1 more fail, and 18 less in HTML5 test), NineSky (3 more fails) are all using stock renderer.

    4. It seems like these all test the CPU for rendering/scripting speed, am I right? In real-world usage though, users are more likely to notice hiccups because of slow disk access (our TFs have super slow NAND for some reason ) or poor UI threading (causing scrolling or touches to not respond). That's one reason why I currently choose Boat Browser over Chrome; Boat feels more responsive even though Chrome renders more accurately.

    5. I had no idea that Dolphin Beta/Jetpack was that good. Thanks for letting me know!
    Tablet: Asus TF101 B9O 16 GB + 32 GB microSD | SOC: 1.4 GHz Tegra 2 | OS: Team EOS 4 JB 4.2.2 (latest nightly) | Kernel: KAT 103 | Dock: B7O
    Phone: Samsung Galaxy Nexus i9250 16 GB | SOC: 1.2 GHz OMAP4 | OS: stock JB 4.1.2 (JZO54K)
    Laptop: Dell XPS 1340 | CPU: Core 2 Duo P8600 w/ 4GB DDR3 dual-channel | OS: Win7 Pro | Other: copper mod

  6. #6
    Jazz
    Member #
    20215
    Join Date
    Jan 2012
    Posts
    30
    Liked
    10 times
    Sorry Zalbee, I didn't inventory my data. Now I wish I had and I could just give it to you guys to play with. So I don't have the results to give standard deviations. I just made a quick mean and jotted it down. Ringmark and HTML5 always gave the same result (except that Ringmark initially gave me a 1 test fail for Dolphin on R.1, which seemed odd to me, so I ran it again several more times, and always got the same result). Of those remaining, I can tell you that by far I saw the greatest deviation with Browsermark, and the least deviation with Kraken and Sunspider. I can only recollect several tests from the latter two benchmarks that listed a possible error of greater than 2%. It was always 1.5% or less. This is why I tiered the results. For example, I found the performance to be so comparable in Tier 3 that I consider the rank within that tier to be nominal. IMO, if you're okay using a less powerful HTML5 browser, then in that Tier, choose based on which layout and features you prefer. On the other hand, even a 5% deviation wouldn't approach closing the gap between Dolphin and those browsers.

    What software would you recommend for graphing this? I have Powerpoint, but that wouldn't embed. Also, Point #4 is a solid. I'd have to read more to understand the relevance of disk access and UI threading to TF101 performance, specifically.

  7. #7
    Starscream
    Member #
    27085
    Join Date
    Feb 2012
    Tablet
    TF101
    Posts
    327
    Liked
    75 times
    No problem on the SD. I think the basic trend is clear, it's just something I wish more benchmarkers would do (most tech sites forget this). I agree with your tiering though, seems like the best way to sort it.

    Excel is pretty easy for graphing. It's also convenient to just enter all your data there, and sort them later (plus it has that STDEV function.)

    The browser on Transformer Prime has been reported to be slow because it caches pages to disk (slow), with Browser2RAM fixing it by caching to RAM instead. How it works

    The threading point was purely speculative. Without source code, we can only guess at what's going on internally, based on subjective feeling of "responsiveness."
    Tablet: Asus TF101 B9O 16 GB + 32 GB microSD | SOC: 1.4 GHz Tegra 2 | OS: Team EOS 4 JB 4.2.2 (latest nightly) | Kernel: KAT 103 | Dock: B7O
    Phone: Samsung Galaxy Nexus i9250 16 GB | SOC: 1.2 GHz OMAP4 | OS: stock JB 4.1.2 (JZO54K)
    Laptop: Dell XPS 1340 | CPU: Core 2 Duo P8600 w/ 4GB DDR3 dual-channel | OS: Win7 Pro | Other: copper mod

  8. #8
    Starscream
    Member #
    38181
    Join Date
    Jun 2012
    Location
    Far East -Canadian style
    Tablet
    TF300T
    Posts
    263
    Liked
    58 times
    Thank you for running and sharing this. It confirms my observations in terms of Dolphin. The one issue I have with Dolphin is its inconsistent response to drop down menus on pages. In terms of speed, it leaves everything else I've tried in the dust.
    Happy owner of a TF300T

  9. #9
    Administrator
    Supporting Member

    Member #
    12453
    Join Date
    Oct 2011
    Location
    Devon, UK
    Tablet
    Other - Android 4.X
    Posts
    10,829
    Liked
    2215 times
    Nice thread & great info. I have linked it from here for future reference...
    http://www.transformerforums.com/for...ur-tablet.html


    PLEASE Search for existing threads before posting a new one. Thanks.

    Your opinion matters. But should you disagree - please try not to be disagreeable

    Forum guide - here ~~ Android Guides - here ~~ Cold boot - here

    Adobe Flash Player & Browser Guide here

    Master Help Guide - here ~~ FAQ malware - here ~~ FAQ e-reading - here
    Android device personal pantheon...
    SGS; Huawei Y300; Motoroloa Razr; CnM Touchpad II; Asus TF101; Lenovo A1; Motorola Xoom 2 ME; Samsung Tab 2 7.0

  10. #10
    Starscream
    Member #
    27085
    Join Date
    Feb 2012
    Tablet
    TF101
    Posts
    327
    Liked
    75 times
    Madmick, I took your data, put it into a spreadsheet, and created some graphs. I will be posting them here shortly. This doesn't change your results in any way, just makes them prettier
    Tablet: Asus TF101 B9O 16 GB + 32 GB microSD | SOC: 1.4 GHz Tegra 2 | OS: Team EOS 4 JB 4.2.2 (latest nightly) | Kernel: KAT 103 | Dock: B7O
    Phone: Samsung Galaxy Nexus i9250 16 GB | SOC: 1.2 GHz OMAP4 | OS: stock JB 4.1.2 (JZO54K)
    Laptop: Dell XPS 1340 | CPU: Core 2 Duo P8600 w/ 4GB DDR3 dual-channel | OS: Win7 Pro | Other: copper mod

 

 
Page 1 of 2 1 2 LastLast

Ads

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •  

Similar Threads

  1. Safe browsers?
    By Digita in forum Transformer Pad 300 General Discussion
    Replies: 5
    Last Post: 06-06-2012, 08:57 AM
  2. web browsers
    By madcutter1223 in forum Asus Transformer (TF101) General Discussions
    Replies: 6
    Last Post: 02-26-2012, 10:10 PM
  3. Trans Prime and/orr Android Browsers
    By jrpbulk in forum Transformer Prime General Discussions
    Replies: 12
    Last Post: 12-22-2011, 09:05 AM
  4. For all android browsers, do they have 'save form data' feature?
    By kohkindachi in forum Asus Transformer (TF101) General Discussions
    Replies: 2
    Last Post: 09-27-2011, 08:06 PM

Search tags for this page

android browser benchmark
,

android browser rank

,

android browser ranking

,
android browser rankings
,

best android browser rank

,

boat browser vs dolphin

,
difference between dolphin and mozilla
,
dolphin beta vs jetpack
,
dolphin jetpack peacekeeper
,
mobile browser ranking
,
ranking browser android
,
sunspider rank
Click on a term to search for related topics.

Tags for this Thread

Powered by vBulletin® Version 4.2.0
Copyright © 2014 vBulletin Solutions, Inc. All rights reserved.
Search Engine Optimization by vBSEO 3.6.1
All times are GMT -6. The time now is 05:50 PM.