Sign in with
Sign up | Sign in

Memory Efficiency

Web Browser Grand Prix 7: Firefox 7, Chrome 14, Opera 11.51
By

We made yet another change to how we run the memory tests. First, we're using the about:memory flag to calculate total memory usage for Chrome. Since Chrome runs each tab in a separate process, there is a slight amount of memory shared between processes. Windows Task Manager can double-count this shared memory. Although such double-counting only ends up amounting to 50 MB or so in a 1+ GB 40-tab load, the about:memory flag is more accurate.

The memory totals of the other Web browsers are still tallied using Windows Task Manager.

Light Load

Internet Explorer remains the lightest Web browser when it comes to the memory footprint of loading a single tab. Chrome 14 drops Google from second down to fifth, though it's still under 75 MB. Meanwhile, Firefox 7 moves Mozilla up from fourth place to second. Safari remains smack in the middle, while Opera moves up to fourth place.

Heavy Load

When the load is changed to a hefty 40 tabs, Firefox 7 doesn't disappoint, shaving about 300 megabytes off the Firefox 6 tally from WBGP6. Safari drops to second place, though it's still comparatively sleek at under three-quarters of a gigabyte. Opera retains third place at over 800 megabytes, and Chrome 14 moves up to fourth place, utilizing just over one gigabyte. Internet Explorer drops to fifth, using the most amount of memory (an astounding 1.3+ GB).

Being the only browser to use less than a half gigabyte under heavy load and the second-place finisher at a light load (less than 50 MB), Firefox 7 is the clear winner. Safari takes second and Opera places third. IE9 and Chrome give us the poorest showings in raw memory usage. IE9's only saving grace is its super-low single-tab total.

Memory Management

We combined the two memory management tests into a single chart. The memory totals were recorded instantly after closing 39 of the 40 tabs, and once more five minutes later.

Chrome is the quickest to immediately release memory back to the operating system, followed by Internet Explorer. Opera comes in an even more distant third, trailed by Safari in fourth.

Surprisingly, the champion of memory usage (Firefox 7) does not give much of its monopolized memory back to Windows right away. After an additional five minutes, Chrome settles in at just 20 MB over its single-tab usage, while IE9 drops back down to just 10 MB over the single-tab tally. Firefox 7 finally gives all but 160 MB back to Windows after a few minutes. While this is still four times the memory tied up in a single tab, it is a good 100 MB less than Firefox 6. After the additional five minutes passes, Safari and Opera both drop down to around 350 MB, placing fourth and fifth (respectively).

Chrome is still a powerhouse when it comes to memory management, needing no time at all to release unused resources back to the operating system. IE9 needs a little more time, but remains incredibly efficient. Firefox 7 gives us the biggest surprise here, using the least amount of memory under a heavy load and dropping to just over 150 MB after five minutes. Safari and Opera are fairly mediocre in this discipline, though neither is as bad as they once were. Comparatively, they're at the bottom of the heap.

Naming a winner for overall memory efficiency is really tough this time, but we have to give the victory to Mozilla. Firefox 7 is simply never weak in our testing.

Display all 131 comments.
This thread is closed for comments
Top Comments
  • 21 Hide
    makaveli316 , September 30, 2011 5:48 AM
    "Until another browser beats Chrome in the speed/performance benchmarks I'm sticking with it."

    lol people still think they can feel the difference in terms of speed in real world performance and there's still people that doesn't use a browser for their needs and preferences, but just because they have seen some silly benchmark.
    Ridiculous. I bet those are the same people that are always complaining in the forums about crashes, viruses and blue screens.
  • 14 Hide
    Anonymous , September 30, 2011 9:30 AM
    This article (as well as the preceding series of articles) is a great example of how amateur journalism can be misleading, or just downright wrong.

    On the surface, everything looks good - the author sets out a methodology, clearly presents the results, and draws conclusions based on them. Unfortunately, in doing so he reveals his severely lacking knowledge of testing methodology, the browsers themselves, as well as how one interprets the results of benchmarks.

    To aggregate across criteria such as "performance" and "standards compliance" (never mind the fact that HTML5 hasn't yet been drawn up), using an arbitrary weighting system, and then conclude that one browser beats other "overall" is nonsensical.

    Nowhere has the author talked about relevance (this is critical) or statistical significance of his tests. I'm sure he put in a lot of effort into the article, and that it was written out of the best of intentions; however, this article remains a jumble of random tests clumsily grouped together. For example, can the author explain to the readers why the removal of SVG fonts in the ACID3 test is important? Should browsers have support for SVG fonts? Should one test for it? If he can't, he's just mechanically running benchmarks that he's found on the internet.

    Obviously it's easier to criticise - but it's much more beneficial for people to actually try the browsers out for themselves (it is free after all) than to read this kind of poorly conducted "showdown".
  • 11 Hide
    killik , September 30, 2011 6:46 AM
    Even better yet,if you use FF7 with the MemoryFox addon,Firefox simply obliterates the competition.try it for yourself.
Other Comments
  • -5 Hide
    gerchokas , September 30, 2011 4:33 AM
    Well done Mozilla!
    Now they could change their famous icon to a more minimalist/modern style and we're done. Speedy AND classy, just like a fire fox.
  • 10 Hide
    shiftmx112 , September 30, 2011 5:04 AM
    Quote:
    Enjoy it while you can Firefox fans


    Indeed. I have been quite content with FF8 though.
  • 8 Hide
    soccerdocks , September 30, 2011 5:04 AM
    Until another browser beats Chrome in the speed/performance benchmarks I'm sticking with it.
  • 0 Hide
    iam2thecrowe , September 30, 2011 5:19 AM
    im sticking with IE, its perfectly fast enough and stable and why should I have to install another browser when it works perfectly fine?
  • 6 Hide
    JOSHSKORN , September 30, 2011 5:25 AM
    I'm on FF10. :D  Nightly 64-bit!
  • 4 Hide
    compton , September 30, 2011 5:37 AM
    I have to say, I do really like chrome. I stopped using Firefox as much one I tried the chrome beta, and now I use IE9 and chrome all the time. I used opera for a while, but Netflix streaming doesn't work with it, nor do many other sites I use.

    Now that IE is good again, I can't fault anyone for using it in lieu of the others.
  • 21 Hide
    makaveli316 , September 30, 2011 5:48 AM
    "Until another browser beats Chrome in the speed/performance benchmarks I'm sticking with it."

    lol people still think they can feel the difference in terms of speed in real world performance and there's still people that doesn't use a browser for their needs and preferences, but just because they have seen some silly benchmark.
    Ridiculous. I bet those are the same people that are always complaining in the forums about crashes, viruses and blue screens.
  • 11 Hide
    killik , September 30, 2011 6:46 AM
    Even better yet,if you use FF7 with the MemoryFox addon,Firefox simply obliterates the competition.try it for yourself.
  • -3 Hide
    killik , September 30, 2011 6:47 AM
    Even better yet,if you use FF7 with the MemoryFox addon,Firefox simply obliterates the competition in the memory management department.try it for yourself.
  • 7 Hide
    frostmachine , September 30, 2011 7:11 AM
    I noticed firefox will use progressively more memory. Even if it's just refreshing the same pages. I use firefox, opera n chrome, keeping them open 24hr/day. It can go from intial 100mb to over 500mb. I don't see this in the other browsers.
  • 5 Hide
    Anonymous , September 30, 2011 7:36 AM
    The Encog Silverlight result is wrong. In the graph IE9 comes third with Firefox fourth, but your conclusions are,
    Firefox 7 shows significant improvement over version 6, moving up to third place. As a result, IE9 drops to fourth.
  • 2 Hide
    Anonymous , September 30, 2011 7:41 AM
    How on earth can Firefox be winner in memory manaegment? Chrome and IE are much better....
  • -6 Hide
    buzznut , September 30, 2011 7:53 AM
    Firefox 7 is supposed to be better at memory management but I am still getting the same messages from AVG telling me Firefox is using too much memory. This only started with 6 so I think they still have work to do here. Maybe if they had longer than 6 weeks between versions...
  • 14 Hide
    Anonymous , September 30, 2011 9:30 AM
    This article (as well as the preceding series of articles) is a great example of how amateur journalism can be misleading, or just downright wrong.

    On the surface, everything looks good - the author sets out a methodology, clearly presents the results, and draws conclusions based on them. Unfortunately, in doing so he reveals his severely lacking knowledge of testing methodology, the browsers themselves, as well as how one interprets the results of benchmarks.

    To aggregate across criteria such as "performance" and "standards compliance" (never mind the fact that HTML5 hasn't yet been drawn up), using an arbitrary weighting system, and then conclude that one browser beats other "overall" is nonsensical.

    Nowhere has the author talked about relevance (this is critical) or statistical significance of his tests. I'm sure he put in a lot of effort into the article, and that it was written out of the best of intentions; however, this article remains a jumble of random tests clumsily grouped together. For example, can the author explain to the readers why the removal of SVG fonts in the ACID3 test is important? Should browsers have support for SVG fonts? Should one test for it? If he can't, he's just mechanically running benchmarks that he's found on the internet.

    Obviously it's easier to criticise - but it's much more beneficial for people to actually try the browsers out for themselves (it is free after all) than to read this kind of poorly conducted "showdown".
  • 1 Hide
    lassik , September 30, 2011 10:07 AM
    FF7 was always my favourite, it had the best storyline. :) 
  • -2 Hide
    Anonymous , September 30, 2011 10:17 AM
    http://blog.mozilla.com/futurereleases/2011/09/30/firefoxbeta8/

    Firefox 8.0 Beta is now available.
  • 7 Hide
    Onus , September 30, 2011 11:16 AM
    Some add-in crash issues in FF a couple months ago forced me to use IE again for a while. Wow, I never realized how many sites had so many ads, and ways to show them! I'm glad FF is more stable again so I can use it once more, although there are still a couple of sites where IE9 is more compatible.
  • -3 Hide
    eddieroolz , September 30, 2011 11:20 AM
    Firefox 7 as the winner of Grand Prix!?! I gotta go check if hell has frozen over right now!

    But on a more serious tone, I honestly thought Chrome had this one again. Looking at the charts my impression was that Firefox never really won anything by significant margins.

    Also, I hope Internet Explorer 10 will arrive soon. My short experience with IE10 under Windows 8 was very pleasant, even better than that of IE9.
Display more comments