Sunspider JavaScript benchmarking

I’ve had a blog post in draft for ages, but this isn’t it. The draft post has extensive tables of stats that lie half-finished, but it was getting a bit out-of-hand, so here’s a very abridged version.

The post looked at benchmarking the JavaScript performance of various browsers. Here is a sneak preview:

Browser Speed and tolerance
Firefox 3 (v3.0.4) 3772.6ms +/- 1.3%
Google Chrome (v0.4.154.29) 1808.0ms +/- 6.6%
Opera (v10.00alpha) 4696.4ms +/- 3.6%
IE6 (v6.0.2900.5512 or something) 47651.0ms +/- 15.1%
Safari (v3.1.2) 4260.8ms +/- 5.1%

Of course these stats mean nothing without some explanation, so here goes.
I’m using the SunSpider JavaScript benchmark. You should read what it’s benchmarking – it does a bunch of core JavaScript stuff, like raytracing and string operations and stuff, no DOM or browser API tests.
I ran these tests under Windows XP, on a Dell Inspiron 9300.

My draft post was/is going to also include the browsers on my MacBook Pro, as well as results from the Mootools Slickspeed JavaScript framework benchmark tool.

The results are interesting, especially with IE6 thrown into the mix. The separation between the old and new generations of browsers is clear.
Opera 10.00 alpha was only released yesterday, running on a new rendering engine, Presto 2.2, which utilizes Opera’s new futhark JavaScript engine, and so I expected it to be a bit faster than it is; Safari, FF3, and Chrome are all faster.

Google Chrome is considerably faster that the others.

I do most of my work on the Mac, and so use Firefox with its extensive array of invaluable extensions, particularly Joe Hewitt’s Firebug, and Chris Pederick’s Web Developer Toolbar. On the PC, I use Firefox, and occasionally Chrome, when I feel the need for speed.

This entry was posted in Commentary. Bookmark the permalink.

3 Responses to Sunspider JavaScript benchmarking

  1. I think this kind of benchmark is a bit pointless because javascript doing raytrace and all is far less common than DOM and others.. I mean it’s not really representative of your typical web page, so it doesn’t really answer the interesting (imo) question, “which browser will display my stuff before the others ?”

  2. Dan says:

    It gives a good indication of the core performance of the engine, which you’d expect would have some bearing on its real-world performance.

    But certainly, the MooTools Slickspeed benchmark gives a much better indicator of actual real-world performance – it runs comparative tests across MooTools, jQuery, YUI, prototype and Dojo.
    I just haven’t got round to finishing my tables of results and write-up for the Mootools stats on my various browsers.

  3. 94Suzette says:

    Hi blogger, i must say you have very interesting content here.
    Your blog should go viral. You need initial traffic boost
    only. How to get it? Search for: Mertiso’s tips go viral

Leave a Reply

Your email address will not be published. Required fields are marked *