Benchmark: Internet Explorer 8 (IE8) vs. Firefox 3.7, 3.6, 3.5 vs. Chrome 4, 3, 2 vs. Opera 10.5, 10.1 vs. Safari 4

By | December 22, 2009 | 29 Comments

Yet again, it’s time to test latest web browser releases, including Opera 10.50.

Tested browsers:
Internet Explorer 8
Firefox 3.5.6
Firefox 3.6 Beta 5
Firefox 3.7 Pre-Alpha 1
Google Chrome 2.0.172.43
Google Chrome 3.0.197.11 Beta
Google Chrome 4.0.266.0 Beta
Opera 10.10
Opera 10.50 Pre-Alpha 1
Safari 4.0.4
WebKit r52483

JavaScript Sunspider benchmark.
Benchmark: Internet Explorer 8 (IE8) vs. Firefox 3.7, 3.6, 3.5 vs. Chrome 4, 3, 2 vs. Opera 10.5, 10.1 vs. Safari 4

Peacekeeper benchmark
Benchmark: Internet Explorer 8 (IE8) vs. Firefox 3.7, 3.6, 3.5 vs. Chrome 4, 3, 2 vs. Opera 10.5, 10.1 vs. Safari 4
* Webkit stuck during benchmark

Overall, Opera 10.5 looks promising, beating Chrome 4 in SunSpider benchmark (with Safari in the lead), end pretty much everyone in Peacekeeper test.

Update: Fixed Opera 10.50 results.

[digg-reddit-me]


About (Author Profile)


Vygantas is a former web designer whose projects are used by companies such as AMD, NVIDIA and departed Westood Studios. Being passionate about software, Vygantas began his journalism carrier back in 2007 when he founded FavBrowser.com. Having said that, he is also an adrenaline junkie who enjoys good books, fitness activities and Forex trading.

  • Dev

    Something is strange because in Peacekeeper test Opera 10.1 and 10.5 have almost identical results. I think you’ve done a mistake somewhere…

    • http://O_o max1c

      indeed there is a mistake big “O” Win
      [IMG=http://www.filedump.net/dumped/1261627786.png]

  • http://thenewobjective.com TNO

    What about dromaeo.com tests?

  • Matheus
    • Ichan

      Damn. Someoen was busy testing all those things.

      • cousin333

        Those are results from multiple people. When you give out that URL, anyone can test. If you test Opera 10.10, and later he also tests Opera 10.10 and has a better machine, then your result will be completelly overwritten. If he uses a browser, that you’ve not used before, the new result will be merged with the formers (like in this case).

  • http://www.favbrowser.com Vygantas Lipskas

    Yes, I have no idea why it reported such results, maybe site was overloaded with hundreds of tests or so… Will try to run it in the morning, when traffic is low.

    • ac

      What were the hardware specs and OS used in the tests? Depending on the processor used you might be affected with one of the known issues, from labs post:

      -No JIT (slow performance) on old processors without SSE2

      • http://www.favbrowser.com Vygantas Lipskas

        Intel Core2Duo E6750
        4GB RAM
        Windows 7 Home Premium x64
        Running on Virtual PC

        • http://www.rachid.nl Rachid

          Running on Virtual PC? How’s that ever going to give you reliable results?

          • Babamomo

            Virtual PC, that explains the results. On my computer Opera 10.5 is 1.17 times faster than Chrome 4 on Sunspider, and Chrome is in turn faster than Safari and latest nightly Webkit.

        • nobody

          virtual pc.. please, do not test it that way, results can get really messy that way. why you use it anyway? you dont want to spoil your main windows installation? or are there other reasons?

          • http://www.favbrowser.com Vygantas Lipskas

            Running on Virtual PC, allows me to test web browsers on default installation.

          • kryz

            …and gives you false results.

  • Daniel Hendrycks

    “Peacekeeper provided strange results for Opera 10.5 (even after running benchmark more than once). As theoretically, it should have scored much higher than Opera 10.10.”

    For me Opera was 4 times faster. Maybe try again. ???

  • http://www.rachid.nl Rachid

    Hi,

    As I was very disappointed with the inconsistencies in the benchmarks here (no offense), I decided to benchmark myself and give insight in what I did and how I did it, following scientific guidelines.

    The results seem very consistent and you can find them here:
    http://www.rachid.nl/browsers/

    Please take a look at it and let me know what you think. I appreciate all feedback. :)

    Of course, whoever wants to, can link to the url mentioned.

    • Krotify

      Very interesting and well done. What about adding Google Chrome 4?

    • cousin333

      Very well done test. I found Chrome 4 not to be faster in JS (or at least, the difference is not significant).

      My short benchmark on Mac OSX Leopard (without the “Snow”) showed (on a Mac Mini), that in SunSpider, Opera 10.5 was first, Chrome beta was second (the difference was about 8-9%, and Safari third, trailing by about 15-16%.

  • Somebody

    I think Opera 10.5 comes out on top in all benchmarks except for Google’s own V8, where it runs it very close. Seems like there is some issue with your tests as all tests run by others seem to agree on this point.

    http://www.betanews.com/article/The-once-and-future-king-Test-build-of-Opera-crushes-Chrome-on-Windows-7/1261519843/2

    http://www.pallab.net/2009/12/22/opera-10-50-released-opera-is-once-again-the-fastest-browser-on-earth/

    http://lifehacker.com/5432054/opera-105-pre+alpha-is-all-about-speed-and-private-browsing

  • http://xpehkto@ya.ru Kirill

    Did you use real pre-alpha (windows build 3172 from http://labs.opera.com/news/2009/12/22/) or leaked build 20192? You can see build number in Help→About Opera.

  • Kirill

    Did you use for test real pre-alpha (build
    3172 from http://labs.opera.com/news/2009/12/22/) or “leaked” build 20192? You can see build number in Help→About Opera.

    • http://www.favbrowser.com Vygantas Lipskas

      Real one.

  • http://www.favbrowser.com Vygantas Lipskas

    Re-run benchmark and got better results, post updated.

    • kryz

      How come all other tests show Opera well ahead of everyone else at SunSpider?

      • Grrblt

        Different computers can get slightly different results.

  • Obs

    These results are incorrect. You can’t run benchmarks in a virtual setup. You should fix it or stop doing it. See znet or betanews instead.

  • Justin

    Tested Both Chrome 4.0 beta and Opera pre-alpha. Chrome on my computer did just a little better at it. Going to test again…

    http://service.futuremark.com/peacekeeper/results.action?key=2cWJ

  • V-2

    I tried Opera 10.50, very impressive results with benchmarks indeed

    then I measured how fast it loads real websites (with webwait com).

    and… rather disappointingly… it was a bit faster than 10.10, true, but still didn’t even beat Firefox (3.5.6), not to mention Safari and Chromium (I use SRWare Iron, based on Chrome 3)

    having said that: well, it’s not even alpha

  • chris

    I’m getting 812ms on Sunspider running under 3.6 RC1 Final.

«
»