Efficiency Benchmarks: Memory Usage And Management
This page ends our performance-based benchmarks, which used to make up the entirety of the WBGP. Now we'll switch gears and look at efficiency; specifically, memory efficiency.
Memory Usage
We open each browser with just one tab, in this case Google.com. We then wait five minutes and record the memory usage numbers from Windows Task Manager. We then open 39 more tabs (the rest of the Quantcast Top 40), wait five more minutes, and again record the memory usage. These first two figures give us an idea how much system memory each Web browser uses, one with a full load of tabs, the other at near idle.
Single Tab
Microsoft's new creation steals the show, consuming only 18 megabytes with a single tab. Previous efficiency champ Chrome drops to second place, using 27 megabytes. Always good for light systems, Firefox only uses 40 MB. Opera ties up just over 50 MB, while Safari hogs almost 100 megabytes.
40 Tabs
Firefox is again the most frugal Web browser for memory usage under heavy load, utilizing only 256 MB to display 40 tabs. Apple's Safari comes in second, employing 800 MB of system RAM, while Opera uses 850 MB. Both Chrome and Internet Explorer 9 use the most amount of system memory to fully load 40 tabs. Chrome uses just shy of one gigabyte, while IE9 uses 1.2 GB.
Memory Management
After we record the memory usage from the 40 tab load, we close 39 of the tabs, leaving only Google.com open. After five minutes, we record the memory usage to see how much system memory the browser gives back. Finally, we wait another five minutes to see if the overall memory usage continues to drop. After closing the browser, we again look at the total system memory to see if any of the applications hang onto or lose memory after being closed. This has not yet happened with any of the browsers tested in the WBGP.
-39 Tabs (5 minutes)
In another serious blow to the previous efficiency king (Google Chrome), IE9 uses less memory after closing all but one of the 40 tabs. Chrome still manages memory extremely efficiently though, holding onto only 42 MB after returning to a single tab. After five minutes, Firefox keeps less than 100 megabytes, a little more than twice the amount of memory before the extra 39 tabs were opened. Fourth-place finisher Safari keeps 330 MB, over three times its original single-tab total. Opera is weak in this area, keeping over 420 MB after closing all but one tab. That's over eight times the single-tab total.
-39 Tabs (10 minutes)
After an additional five minutes, Internet Explorer 9, Chrome, and Firefox remain in the same order. Safari also remains the same, though Opera drops down to 330 MB, beating Safari for the number four spot. At 330 MB, Opera still holds six times the memory of its single-tab load. Safari keeps three times its single-tab load, while Firefox doubles its usage.
Only Chrome and IE9 take memory usage down to less than twice their original one-tab totals, which are already the lowest of the bunch.
Related Stories :
- Web Browser Grand Prix 7: Firefox 7, Chrome 14, Opera 11.51
- Web Browser Grand Prix 5: Opera 11.50, Firefox 5, And Chrome 12
- Google Chrome Pushes Past Microsoft's Internet Explorer
- Web Browser Grand Prix 9: Chrome 17, Firefox 10, And Ubuntu
- Google Upgrades Internet Explorer With Chrome
- Internet Explorer 10 Now Available for Windows 7
- Internet Explorer 10 Preview Coming to Windows 7 Soon
- Internet Explorer 10 to Have 'Do Not Track' as Default
- Next 3DMark to Support DirectX 11, 10 and 9
- Microsoft Removes Legacy Baggage From Internet Explorer 10
I wonder how much microsoft paid tomshardware not to include Firefox 4 RC1
Now everyone is going to switch from Chrome to IE?
Nice comparison Adam. I think people don't give MS credit for IE9 because of their history with web standards and lack of competition with updating their browser since they destroyed Netscape so many years ago. Hopefully they have their head in the game and continue with a somewhat constant update cycle. Currently the transition from IE8 to IE9 was somewhat FireFox'ish with speed instead of waiting for another revision of the OS. The browser definitely surprised me, some quirks here and there that are definitely noticeable and the Chrome/FF influence is obviously present. I wish them luck as the competition heats up.
Will there be a " Web Browser Grand Prix 4: Firefox 4 Enters The Race " after Firefox 4's release (22 Mar)?
i felt the speed once i switched from firefox to chrome and now chrome to IE..
but there is no mention of any addon/extentions for IE..
i mean there is no comparision on who's got nice addons and startup time "with" addons.. since users tend to use web browsers with addons knowingly or un-knowingly..
it would be better if you include those two comparisions as well.. and who takes advantage of GPU and all..
Won't take that long for IE to lag behind.
Won't take that long for IE to lag behind.
lol
maybe not!!
and FFS why doesnt that "submit my comment" button not working in IE!! I am using chrome for that ;(
I like how this article was released one day before Firefox 4 is released.
duck it, i stick to firefox.
Is there a wrong picture on the efficiency page? The 40tabs graph comes up as GUIMark 2?
The 40 Tabs chart is wrong for Efficiency Benchmarks: Memory Usage And Management it shows flash vector charting.
I wonder how much microsoft paid tomshardware not to include Firefox 4 RC1
Then they would have to include every random nightly build of every browser? Comparing the latest release is fair you'll always get one browser or two that is near a new stable version but that's just how it works.
Won't take that long for IE to lag behind.
depends in Microsoft moves up an update schedule that doesn't just include making the browser more secure and fixing bugs. IE8 was pretty snappy back when it was first out it got old in like 3 months compared to other browsers though.
I wonder how much microsoft paid tomshardware not to include Firefox 4 RC1
We only test final products in the WBGP.
Will there be a " Web Browser Grand Prix 4: Firefox 4 Enters The Race " after Firefox 4's release (22 Mar)?
I like how this article was released one day before Firefox 4 is released.
Where are ya'll getting a set date from?
i felt the speed once i switched from firefox to chrome and now chrome to IE..but there is no mention of any addon/extentions for IE.. i mean there is no comparision on who's got nice addons and startup time "with" addons.. since users tend to use web browsers with addons knowingly or un-knowingly..it would be better if you include those two comparisions as well.. and who takes advantage of GPU and all..
We add/modify tests every time with the WBGP. We're already looking into those suggestions.
Is there a wrong picture on the efficiency page? The 40tabs graph comes up as GUIMark 2?
Doh! Fixing that now.
Where are ya'll getting a set date from?
http://groups.google.com/group/moz [...] 9e9650b1b3
Wait a second. Did they really include Safari? Who uses Safari for Windows?
(I'm assuming they tested the Windows version of Safari, because why would you compare browsers for different systems?)
I'm sticking with Opera on my Windows machine and Safari on my Mac.
To be honest I did not read this article, because you're using the "latest and greatest" browsers, but somehow jumped a day early and didn't wait for FF4. Maybe you mentioned it in the article, but if not, then was it really so bad to wait an additional 1-2 days to do this?
... don't know... but this all is not my experience... why there is no firefox 4.0 RC2 ????
Where are ya'll getting a set date from?
http://www.theregister.co.uk/2011/ [...] _22_march/
Although most websites link to this website, so perhaps it was nothing more than a rumor.
Wait a second. Did they really include Safari? Who uses Safari for Windows?
(I'm assuming they tested the Windows version of Safari, because why would you compare browsers for different systems?)
I'm sticking with Opera on my Windows machine and Safari on my Mac.
Opera sucks balls in memory management and lacks handy add-on such as Tab Mix Plus (which is same for Chrome). I would definitely move to Opera if it has these 2 issues fixed.
https://wiki.mozilla.org/Firefox/4/Beta
This article looks like a case of premature e.. something.