Battle of the Web Browsers - Who Uses Less Memory?

13
Mar 2010

Battle of the Web Browsers - Who Uses Less Memory?

comment icon7 comment(s) |

Within our group of web developers, we regularly review how a website will render across all web browsers to ensure everyone gets an equal experience. Of course, everyone has their favorites and everyone enjoys complaining about all the others.

At work, I typically use Chrome or Firefox as I like the simplicity of Chrome and the add-ons of Firefox...so it just depends on what Im doing at the time. However, my work PC is fairly powerful so I do not have to consider the 'weight' of a browser.

This is different at home.

Our desktop computer that sits in our home office is older than my kids - I bought this beast to move into the new millenium. That said, it is a work horse that simply will not die...and I am someone that cannot throw away anything that works, even if it is seriously limping.

This brings me to my post: What browser is the lightest, using the least amount of precious memory as it does its basic functions? I was surprised with what I found.

The Tests

To test this, I decided to use two scenarios:

  1. Open browser, go to www.google.com.  Record the memory usage.
  2. Open browser, go to www.facebook.com.  Log in.  Record the memory usage.

The idea was to test the most basic operation first, then test with slightly more memory being burned through intense javascript use.

The Contenders

I chose the major players in the browser game, knowing that some of the others arent known for being lighter and generally command less mainstream respect anyway.  I used:

  1. Firefox 3.5.08
  2. Firefox 3.6
  3. Internet Explorer 8.0.6001
  4. Chrome 5.0.342 Dev
  5. K-Meleon 1.5.4

Those of you unfamiliar with K-Meleon should be told in advance that it will win this test, hands down.  If uyou have an old, slow machine that runs Windows - you should run K-Meleon.  It uses the same layout engine as Firefox (Gecko), but uses the native Windows API for the user interface.  This results in a much lighter browser as it doesnt require its own user 'weight'.

Off To The Races

So, with all the contenders closed - the battle began.  Here were the results:

Browser Viewing Google  Viewing Facebook 
FF 3.5.08 42,344k 66,531k
FF 3.6 37,636k 60,688k
Chrome 5.0  36,280k 65,964k
IE 8.0  27,892k 58,316k
K-Meleon 21,648k 33,316k

Of the 'mainstream' browsers, Internet Explorer was the clear winner.  Unexpected.  Of course, an argument could be made that IE uses some libraries or resources that are already running under Windows - but Im jsut comparing the data for entertainment purposes (sad, I know).

Interesting Notes

A few interesting things were found that I didnt expect, and that I am sure someone will chime in on:

  1. In Chrome and IE, there was only one instance of the browser running and only a single tab - but once I had entered the Facebook URL, a seperate 'image' appeared in the Windows Task Manager.  This essentially doubled the amount of memory used.
  2. In Firefox 3.6, it opens and loads Google at 37, 636k; but then slowly grows in memory usage as along as I left it open.  After a couple of minutes, seemed to level off at around 67,000k.  This will probably be fixed in the next release of FF as 3.6 just came out, but always interesting to note.

Conclusion

Firefox, the cult-followed leader in the mainstream bracket, is the unrivaled hog. 

IE has gotten significantly better at handling itself in the real world.  No more is IE the resource sucking hog just waiting to devour whatever you fed her.

Chrome looks good, but still isnt as light as IE in basic work.  However, Id guess that a resource intensive web application would still run best under this browser.

Those conclusions noted, I think the main thing we learned today is that Dalton had some free time this morning that probably could have been better spent in any number of endeavors.  Next time, I'm going to ride my bicycle or blog about Twitter.

Comments

March 15, 2010

eshiarla

Noticed the same over the last year

As a result Firefox has turned into strictly a development browser for me. In my experience these results are even more heavily weighted against Firefox once any multimedia site is loaded up (Hulu.com, Last.fm, TheSixtyOne.com, etc.).

March 15, 2010

Cole Shelton

Firefox is dead to me. I was

Firefox is dead to me. I was excited about TraceMonkey, but the speed improvement was overshadowed by FF's continued resource hogging. The only time I open it up is to use Firebug, as Eric pointed out, but since I do little development, that is not often.

March 18, 2011

Steve

Browsers

Sorry don't believe these results at all. IE 8 is a horrendous system hog that even shows up twice in task manager. It ramps up memory during usage to the point that it stops working entirely.

September 30, 2011

Anonymous

BROWSERS

I like Firefox but it is really a system hog and IE does show up twice in task manager at execution. It is annoying while using 3 windows the system advises to close and re-open IE due to a significant use of memory.

January 31, 2012

Anonymous

I'm afraid I'm going to have

I'm afraid I'm going to have to go with K-Meleon. My machine is a Toshiba Sat from 2006 and I'm running google chrome, in fact I love google chrome. But what I don't like is whenever I'm trying to play an app game on facebook, shockwave stops responding for about 3 to 4 min. Then when it finally clears, and I click to do something else like tab over or go to scroll up, it freezes again for another 2 min. I can't play any games for it freezing. Especially games like bingo. When it finally starts responding, the games either over or the caller has called 9 or 10 more numbers and you're desperately playing catch up. Bugs the crap out of me. So, gonna try K-Meleon. Thanks for the info.

July 6, 2012

The Rob 2.28

What About Safari

We are having a remote in performance problem at work, and so I have been trying to dissolve this issue by choosing a lighter weight browser. I wanted to mention there are many other browsers out there that you didn't try out. Safari, Opera being two. I was doing the same tests as you, and perhaps our facebook pages are a little different but I also noticed that even K-Melon is somewhat based on the amount of resources available. At unlimited resourced I found these results:

Firefox 145,000
iexplorer 90,000
k-melon 416,000 k
navigator 414,000 k
opera 91,000 k
safari 62,000 k
chrome (multiple instances) running at 180,000 k

I would also like to note that there may be other services started that I have not included in this instance, and I am looking into which svchost.exe might be hiding form the normal view for the resource monitor for safari because right now it looks like the best bet. If anyone has the time, this performance test really should be done with different Virtual Machines set up the same exact way with benchmarking and memory usage with each browser installed on a separate virtual machine.

November 14, 2012

Anonymous

Chomed piggy...

I have an issue with the data being put forth towards Google Chrome.

This test is single tab single window.

Unlike most other browsers that i'm aware of,
chrome loads a set of processes for each tab.

in my case. I've got 9 processes for 5 tabs.

this will most likely shoot chrome to the very top of the hog list.
I've not tested this on the fox or exploder but i'm told by friends
that they don't experience this type of behavior on the other
two mainstream browsers.

Might be worth another look see no?

Search