Monday, September 29, 2008

Browser Speed Tests: The Compiled, Up-to-Date Results [Performance Tests]

Source: http://feeds.gawker.com/~r/lifehacker/full/~3/404025740/browser-speed-tests-the-compiled-up+to+date-results

Back when Firefox 3's final release candidate dropped, we ran some tests to compare its page-loading, memory use, and technical timing to Internet Explorer 7, Opera, and Safari for Windows. Then Google Chrome arrived, so we pitted it against the betas for Firefox 3.1 and Internet Explorer 8, and shared the results. The tests were by and large the same, but many commenters wisely asked to see all the results, betas or no betas. Well, today we've patched together all our data, thrown in a fresh test of the Opera 9.6 beta, and we're sharing all the graphy goodness. Read on to see a full comparison of the major browsers you can load on Windows.

The tests

You can read up on our testing methods in their entirety at the original testing feature. I've come to realize, however, that between all three rounds of testing involved, the "8-page load" test may be the most vulnerable to variables—some of the pages included are quite dynamic, so if, say, Gizmodo puts up a large number of videos or huge pictures, it could affect the total loading time. Other than that, though, the page-loading tests are run by a human watching a timer, the JavaScript from Sean Patrick Kane's web test, the CSS from a downloadable form, and memory use from checking Windows Vista's Task Manager.

Test 1: Page Loading—Winner: Opera (9.5)!

No surprise that Opera 9.6's beta performed just as well as the official release, on start-ups both both cold (right after b! oot-up) and warm (having launched at least once). I'm heartened to see comparable results between the first batch of browsers I tested and their newer betas.

The next speed test, loading eight pages from a bookmark folder, left me scratching my head. Why did the newer betas take so much longer to load a similar set of pages? As stated above, my best guess is the dynamic nature of at least one page in the group, but Opera was tested separately from the other betas, and didn't gain much in speed.

Test 2: JavaScript & CSS—Winner: Safari! (by a nose)

It's hard to beat Safari's performance in both Cascading Style Sheets and the JavaScript code that fronts so many webapps. It has to be noted, however, that most browsers, other than Internet Explorer, don't out-run one another by a huge stretch in JavaScript; Chrome and Safari, though, pull ahead on CSS.

Test 3: Memory Use—Winner: Firefox!

It's reassuring that Mozilla puts so much effort into memory usage in Firefox 3 releases—seeing as how most readers of this site are more than open to extension suggestions.

So that's all the testing data we have on the latest web browse! rs here at Lifehacker Labs. Got another set of test results you put faith in? Surprised at any of our outcomes? Tell us about it in the comments.