Posted on 06/22/2011 6:26:22 AM PDT by ShadowAce
Officially, Firefox 5 is scheduled for release tomorrow. However, users of the beta channel have already got their hands on it. One of the promises for Firefox 5 is better performance. We took tested Firefox 5 and benchmarked it against two other browsers - Google Chrome and Opera. We used the latest version of Google Chrome from the beta channel and the latest stable version of Opera - Opera 11.11.
The SunSpider JavaScript Benchmark is a test for the JavaScript engine of the browser. In this benchmark, lower scores (time) means better performance. In our earlier benchmarks, Firefox is always slower than other browsers such as Chrome and Opera. Let us see how well Firefox 5 does.
As you can see from the graph above, Firefox 5 has made huge improvements in its JavaScript engine. It is now much faster than both Chrome and Opera. However, the difference is very small - just 30ms faster than Chrome and 8ms faster than Opera. So, finally Firefox can no longer be considered a slow browser.
Peacekeeper is yet another highly regarded test for the JavaScript engine. Unlike the SunSpider test, Peacekeeper tests the normal JavaScript applications such as animations, navigation, forms etc. In this test, higher scores means better performance. Here is the result:
The result from the Peacekeeper test is very different from that of the SunSpider test. Here, Opera got the highest score while Firefox 5, which did best in SunSpider, performed the worst of the three browsers.
The HTML5 Test is a test to determine how many of the HTML5 and related specifications are supported by the browser. It does not test how well the specifications are supported. Scores are awarded for support of the HTML5 standards and bonus points are awarded for supporting related drafts and specifications. Higher scores are preferred in this test.
In this test, Chrome takes the lead by scoring more than 300 points. Opera and Firefox are neck to neck with Firefox scoring 254 and Opera scoring 255. The performance of Firefox 5, is a huge improvement from that of Firefox 4 which manages a score of just above 200.
This is the time it takes for a browser to start on a machine that has been just restarted. Once the machine has finished loading everything, the browser is restarted and the time it takes for it to load a locally stored file is found out. Obviously, lower cold start time is preferred. Here is how the browsers performed:
Opera takes the longest time to start here at around seven seconds. The other two browsers have almost the same cold start time at a little above four seconds.
A test for warm start was not done because all of the three browsers have a warm start time of less than a second and it was impossible to time it properly.
The different browsers takes up varying amount of RAM. To test how much each of them is taking up, we loaded 25 popular websites and check the memory consumption of the browser after opening every five websites. Extensions, add-ons etc. which are not installed by default are removed. Here is the result:
The result here is staggering. While Chrome takes up more than 1GB of RAM on opening 25 websites, Opera and Firefox takes up a little more than only 400 MB. Firefox and Opera takes up nearly identical RAM throughout, although Firefox takes up a little bit less than Opera.
As you can see, Firefox 5 does very well in four out of the five tests. Here are some of the main points about Firefox 5 from the tests:
Oh, OK, I see it only opens when in a posting window. What threw me off, is it claims to be available on any web page.
You can add a Xinha button to any menu bar, which will pop up the editor window at any time.
I’ve had the button on my menu bar in the past, but never used it, I just use it for FR and other forum postings.
Searched for “mapmonkey” and couldn’t find it......Do you have the address? Thanks
Huh. I’ve added it to my menu bar, but it grays out. Even when I have this comments window open it’s grayed out. The only way I can use it is with a right-mouse-button click.
After being open for five or six hours and used intermittently, Firefox 5.0 locked up on me and it took the usual many attempts to close it,”Program not respondng” and several attempts in Windows Task Manager.
Some improvements, but that part isn’t corrected.
I’ve been using it for about the same amount of time. I’ve added a few add-ons, and have used it continuously. No problems. I can’t say FR is much (if any) faster, but Firefox does seem to boot faster. I haven’t had to reboot my PC during this time, so I can’t say a fresh Firefox boot will be any faster.
bfl
“So how do you prevent the memory leaks?”
You can’t. Have tried every thing.
So I just kill the FF process every 15 minutes and
start over. And I mean kill it, because just closing
FF leaves it in for 30 seconds or more while it gives up all
its memory. I just use the session restore feature after
I restart it after the kill.
I think it’s definitely faster and improved. I go to a lot of different websites and load other things into memory so there’s was a lot of opening and closing of websites and other applications before it locked up.
Placemark.
Downloaded Firefox 5 for Fedora. It seems just about as sluggish as version 4, perhaps a tick faster, but hard to tell. Chrome still seems much faster on Fedora.
FF has no such memory leaks and hasn't had for several major versions. Add(s)-on do.
I can run FF 4.01 and 5.0 on an old laptop with 192 MB memory and it's fine. Generally around 128 MB in use + caching, buffering. Very little use of the swap file.
I don’t think Chrome on Linux has extension-accessible timers implemented on the Linux version unless they were just added in Version 12. That was always the killer for me. Plus it seems to need a little more (not much) than the 192 MB on the laptop mentioned above. It is fine on another laptop with 384 MB.
I am about as far from a techie as anyone could be and still have electricity...
I do run FireFox and hope to learn a tiny amount on this thread - brain can’t absorb much though.
Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.