New Browser Test Shows Safari 4.0 is the Fastest
The benchmark is designed to test the performance of Internet Explorer, Firefox, Safari, Chrome, and Opera. It is completely online-based and there is nothing that needs to be downloaded. (Rather, nothing needs to be downloaded to run the benchmark itself; an optional system scan that collects data about your hardware requires the installation of an ActiveX app.) To run it, all a user needs to do it click on the "Benchmark Your Browser" button on the Peacekeeper webpage.
The last two tests, however, do have some visual elements to them. Peacekeeper's fourth test, "Rendering," measures a browser's "ability to render and modify HTML elements, and its ability to show, scale and animate images." The final test, "Real life example," combines elements of the first four tests to emulate "typical webpage function, such as verifying checksums, loading, sorting and searching for data."
Without performing a system scan (the optional ActiveX app), it took Peacekeeper about 160 seconds to run in Internet Explorer 7.0 (IE7) on our HP Pavilion Elite m9550f test system (2.5GHz Intel Core 2 Quad Q9300, 8GB DDR2 SDRAM) running Windows Vista Home Premium SP1 (64-bit). On our test system, IE7 garnered a score of 177. Based on the scores that others are reporting from running Peacekeeper on their systems, IE7 is the slowest browser. Safari 4.0 consistently gets the highest scores, followed by Chrome.
With new browser-on-the-block Google Chrome 1.0, as well as waiting-in-the-wings browsers, Microsoft's Internet Explorer 8.0 (IE8) and Apple's Safari 4.0, the browser wars are just starting to heat up. This week, Microsoft made claims that IE8 is the true performance leader of browsers. This claim comes just two weeks after Apple made a similar statement about Safari 4.0. Peacekeeper aims to settle these and other similar disputes with a set test methodology applied across all browser platforms. Contrary to the implied resolution-settling moniker of this new benchmark, however, this new test is likely to just add fuel to the fire as some vendors will use its results to tout their browser's prowess, while other vendors will likely decry a faulty methodology. Perhaps our biggest fear is that some vendors will tweak their browser's performance to do well on the Peacekeeper benchmark, with optimizations that have little to do with real-world performance. This might sound contrite, but we have seen similar behavior before from vendors when it comes to benchmarks.