We've done a good number of these tests now, and the methodology remains much the same here—testing how long it takes for browsers to start up and load pages, and how much memory is eaten up, from a user's perspective. We don't use a fancy multi-protocol benchmarking suite, mostly because each suite is subjective to a developers' preferences and recording errors.
Browser start-ups are measured from double-clicking to the load of a locally saved Google home page, as "cold," or right after reboot and "warm," with the browser already having run once. Each browser is given a folder full of nine sites—up from eight in previous tests—and forced to load them all at once. Those timings are measured with Rob Keir's timer app, and done three times each and averaged, with way-out results discarded under the assumption of general computer wonkiness.
The memory results come from Windows' own Task Manager. In the case of browsers that use multiple processes, a screenshot is taken of the processes screen after the browser "settles," and a sum is added up there. In Chrome's case, we used the about:memory report, as Chrome/Chromium reports its memory a bit differently. The memory used by the about:memory tab itself is subtracted from the total reported memory.
Finally, these tests were done on a Lenovo ThinkPad T61p, with 2GB of RAM and a 2.0 GHz Centrino Duo processor. For the first time since we started these tests, they were also performed on a fully updated, freshly installed final edition of Windows 7 Home Premium, rather than a fresh copy of Windows XP Professional. If you wanted to see how these browsers fared on Windows 7 versus XP, our last batch of speed tests compiled aggregate results for all browsers, with the same exact tests, minus the specific Chrome memory measurements.
Enough with the rules and regulations—here's how the browsers fared in our latest round of slightly obsessive-compulsive ratings. Note: Click any of the images below for a much bigger, clearer view.
Boot-up and page loading—Winner: Google Chrome (Development/Beta)!
In our last go-round, Opera's final 10th edition was just on the edge of beating Chrome's "stable" 2.0 release at cold and warm start-up times, while Firefox 3.5 tried its gosh-darn best. Either Chrome's development version made some huge changes between 4.0.203 and 4.0.223, or adapted better to Windows 7, or Chrome 2 bulked up a bit, because Chrome Dev consistently started up faster than its stable brethren. Meanwhile, Firefox 3.6 beta 1 bore out its claims to boosting start-up times, being nearly neck-and-neck with Chrome's stable edition, and Opera continues to be an impressively snappy browser.
When it came to loading nine tabs at once—all the browsers' home pages, plus Google.com, YouTube, Lifehacker, and Gizmodo—Firefox 3.6 beta 1 continues to almost line up with Chrome's stable release, while Chrome development took the prize. As with our last tests, Internet Explorer 8 was surprisingly quicker at multi-page loading than Safari 4 or Opera, and, truth be told, faster than Firefox 3.5 and nearly on-par with Firefox 3.6 or Chrome 2. It's installed and semi-running from start-up, so you might think it has an inherent advantage in the newest Windows release—but, then again, look at how long it takes to start up.
JavaScript—Winner: Google Chrome! (For Now)
As mentioned above, we're planning to implement DOM/query-based testing into this test to give a fuller picture of how browsers manipulate code and move page elements around on the fly. On a straight-up test of JavaScript prowess on Mozilla's testing suite, though, both of Chrome's versions seem far ahead, Safari earns a sure second place, and the rest are fairly matched (Internet Explorer 8 refuses to finish on Dromaeo, so we left it out of this chart).
We'll note here, as many browser developers have, that pure JavaScript speed tests can be considered akin to horsepower engine tests—they don't tell you everything about performance, and might be adjusted for by browsers' JavaScript engines.
Memory use—Winner: Firefox 3.6 beta 1!
Firefox 3.6 beta 1 continues Mozilla's mission of reigning in memory use, both at start-up and with content loaded, with even slimmer returns than 3.5. Opera 10.01 was a little slimmer than Safari this time around on Windows 7, and Chrome, measured with its own memory counter, still takes a fairly hefty commitment to pull off that speed. Then again, if you're rocking a laptop with 4GB memory, that might not matter so much.
Grading the results
We're going to try something a little new here, assigning a number grade to each browser's performance in the various categories. They'll be based on the raw numbers, but the ranking assigned to them is our own doing.
Given seven browsers competing in each test, we'll rank each browser's performance on a scale of 1 to 7, 7 being the best performance, 1 a marked worst, and awarding a tie when the results are close enough to be affected by one editor's timing reflexes. The points for each test were added together to determine the final score for each browser. Here's how they fared in each test, to our eyes:
Scores (out of 28 possible)
- Google Chrome 4.0.223.11: 23
- Firefox 3.6 beta 1: 21
- Google Chrome 2 (stable): 19
- Firefox 3.5.4: 17
- Safari 4.03: 17
- Opera 10.01: 16
- Internet Explorer 8.0.7600: 13
Conclusive, scientific results? Not so much. That said, we're glad to see measurable improvement in Firefox 3.6 beta 1, and we're always eager to see what Opera, Chrome, and Safari have been building into their latest releases. Internet Explorer, hey, you just keep keepin' on, yeah?
Send an email to Kevin Purdy, the author of this post, at kevin@lifehacker.com.
Nenhum comentário:
Postar um comentário