Performance of 10 JavaScript Browsers Using Peacekeeper

[Mastermind]

We Have talked yesterday of the release of a new benchmark tool for browsers launched by Futuremark: Peacekeeper. This tool was created to measure the performance of JavaScript browsers, and only them. The result does not take into account the general speed of page display, or various criteria such as the W3C compatibility. Only pure JavaScript performance are reflected in a series of tests on common sites like Facebook or Gmail.

We therefore wanted to know which browser is doing well under Windows on Peacekeeper. We selected the five most commonly used products, each in two versions: a stable, and the latest development version public. Software testing are:

• Internet Explorer 7
• Internet Explorer 8 Release Candidate 1
• Opera 9.64
• Opera 10 Alpha
• Firefox 3.0.7
• Firefox 3.1 Beta 3
• Chrome 1.0.154.42
• Chrome 2.0.169.1 Beta
• Safari 3.2.2
• Safari 4.0 Beta

The test machine is equipped with a Core 2 Duo P8600 (2.4 GHz) and equipped with Vista running Service Pack 1 and all the latest updates. Here are the scores that we obtained, from largest to smallest:

• Safari 4.0 Beta: 1267
• Chrome 2.0.169.1: 1029
• Chrome 1.0.154.48: 1017
• Opera 10 alpha: 861
• Safari 3.2.2: 829
• Firefox 3.1 beta 3: 668
• Opera 9.64: 545
• Firefox 3.0.7: 469
• Internet Explorer 8 RC1: 333
• Internet Explorer 7: 195

All that can be drawn for the moment is that work on Apple's Safari JavaScript engine has borne fruit. The beta of Safari 4 crushes in on the competition, with the exception of the two versions of Chrome are not that far away.

It is important to note that JavaScript performance do not represent the performance. This is an indicator, but those who have tested the example build 7048 of Windows 7 have seen an impressive acceleration of the reactivity of Internet Explorer 8. Yet it ends penultimate performance on JavaScript.

The perceived speed can not be determined by a benchmark. Indeed, among a crowd of browsers that offer performance deemed "equivalent" by the majority of users, it is especially the features that make the difference. For example, some prefer a Firefox 3.1 with extensions that Safari 4.0 is twice as fast on JavaScript, but naked as a worm.

Note: If you Have Also Tested The Browsers please post your Benchmarks Results Here

[WarHammer]

Hello , The problem is that there are so many parameters that make a test "reliable" seems rather risky ... With or without Javascript? Simple HTML or CSS 2? GIF or PNG? With what bandwidth (Opera Mobile is better in "low BP" how to deal with bugs in IE ...? It is better a set of indicators to optimize rather than a "general test" ...

[Sloane]

Hello , I want to be nasty, but safari and chrome on their Firebug. The Safari is perhaps even better than Firefox (as against it is not extensible with modules). it will been interesting to test the latest version of firefox 2. Also , since I went to 3, I found what slowdown it slef . Javascript may not be involved, but it will provide an answer.Thanks

[Rodney123]

Hello , anyone know what score Does IE6 get ? and the difference in performance between the 32 and 64bit versions of browsers ?
However, this benchmark is not at all representative of the sensation of speed of a web browser to use (as opposed to bench 3dmark who represented much more effective use of the system, quite similar to a video game )

According to Microsoft under IE8 processor time used by javascript represents less than 4% of CPU time used to load a web page (average performed on the 100 most visited sites in the world) so even if microsoft 1000 multiplied by the performance of its JavaScript engine, can not feel any difference in performance on most sites rather it is the side of the computing time of the page layout and rendering the IE team has looked primarily to improve the speed of surfing with IE8