Microsoft has published a document (video here) which examines the issues involved in carrying out benchmarking and performance analysis on browsers. Contained in this document is a chart which suggests that IE8 is faster at rendering many of the Internet's top 25 website (as measured by comScore) than either either Mozilla's Firefox or Google's Chrome.
Here's the chart:
Out of the 25 sites listed, IE8 was the fastest at rendering twelve of the site, while Chrome won on nine while Firefox 3.0.5 was first across the finish line for only four of the sites. So how does this fit in with all the testing which seems to suggest that both Firefox and Chrome as faster (much faster) than IE8?
Well, while I can't flaw the testing process (which does seem quite thorough), I do feel that the choice of test pages is pretty poor. Does Google's landing page really give the browser much of a workout? Yahoo!? Wikipedia? Sure, these landing pages are very, very popular, but most are very tight and highly optimized. What I'd like to see is how the browsers compare when you're looking at a results page on Google or a really big entry on Wikipedia or when looking at the store on Apple's site. If you're just visiting the landing sites of these sites, then IE8 will be faster about half the time, what I'd really like to see is how it performs as you're browsing away from the landing page and actually start doing something on the site you're visiting.
So, while I'm reasonably happy with the way the test was done, I feel that the choice of sites was poor and don't reflect well what people do with their browsers. Also, my experience with the different browsers seems to support my belief that IE is not the fastest for general browsing.