Tom's Hardware Pits Newest Firefox, Opera and Chrome Against Each Other

An anonymous reader writes "Firefox 7 was released a couple days ago, and now the latest Web browser performance numbers are in. This article is the same series that ran benchmarks on Mac OS X Lion last month. This time around the new Mozilla release is going against Chrome 14 and Opera 11.51 in 40+ different tests on Windows 7. Testing comes from every category of Web browsing performance I can think of: startup time, page load time, JS, CSS, DOM, HTML5, Flash, hardware acceleration, WebGL, Java, Silverlight, reliable page loads, memory usage/management, and standards conformance. The article also has a little feature on the Futuremark Peacekeeper browser benchmark. An open beta of the next revision has just been made public. This new version adds HTML5, video codecs, and WebGL tests to the benchmark. It's also designed to run on any browser/OS/device combination — e.g. Windows desktop, iPad, Droid 2, MacBook, Linux flavors, etc. Another great read, a must for Web browser fanatics!"

Comments

Re:Is performance really an issue?

[timeOday]

It all depends on the hardware - I just got an old P4 for my kids to use, and the web suddenly seems surprisingly bloated and slow.

Re:This is one of the worse bench compil ever

[abhi_beckert]

I'll mention two "wrong" items I spotted:

Light memory usage (1 tab):
firefox: 438MB
chrome: 134MB
conclusion page: "strong" for firefox and "acceptable" for chrome

memory management (after closing 40 tabs):
  * firefox: 438MB immediately after, 161MB five minutes later
  * chrome: 134MB immediately after, 94MB five minutes later
conclusion page: "winner" for firefox, and "strong" for chrome.

WTF?!

Re:This is one of the worse bench compil ever

[Baloroth]

Where are you getting 438MB for Firefox 1 tab? The value listed is 42.3MB, and 475.3MB for 40 pages. I agree about the memory management thing (kinda). Firefox probably caches the pages to reload in case you open them immediately, though, so the fact it unloads that memory later but not immediately might be counting for it... IDK.

Stability Tests

[ohnocitizen]

I'd love to see a multi-platform (where possible) stability benchmark across the major browsers:

Opening the same site in 10 tabs. in 100. At what point does the browser crash? What is the memory usage?
Now open the same youtube video in 10 tabs. In 100. Repeat the above.
Do the same with trailers.apple.com.

Next, open a youtube video in 10 tabs for each browser, and log how long that pid remains active. Is it still there after a day? After a week? Or does it crash with no user interaction?

I wonder where Firefox would stand in the ranks after tests like the above.

Re:Will this finally shut the trolls up?

[AK Marc]

The problem is that people compare performance of add-on-free Firefox against the others, then compare features as if every possible add-on were installed and working perfectly with no decrease in performance. That's why I like Opera. It's like Firefox with the add-ons I like, but I can leave the add-ons on and not take a performance hit. Given the performance/features of Opera, they beat Firefox for every test that isn't contrived by Firefox fanatics where they are similarly featured.

Re:Is performance really an issue?

[Anonymous Coward]

Sort out H.264 support. (I don't care if it's not free-as-in-whatever

You may not, but the Firefox developers do, and they care because Firefox is a free product and they're the ones who will be sued by MPEG-LA.

it's a much better format

Let's all pretend you can notice the difference between H.264 and WebM at the resolutions and bitrates employed for streaming on the Internet.