Safari vs Firefox vs Chrome on Mac: Chrome Wins!
Almost a month back, I had done a comparison of javascript performance on the latest available betas of various Mac OS X browsers. Chrome edged out Safari narrowly in the Sunspider test and was a clear winner in V8. Chrome has now widened the gap and continue to lead in JavaScript performance. Opera has been dropped from the comparison because it was no match for the other browsers.
I love Firefox. It is still my main browser. But Firefox is slow compared to chrome and safari. It takes some time to start up (not a big problem), uses more CPU (not good especially on a laptop), freezes sometimes and in general is not snappy. I have been using Firefox from the time it was called phoenix and it was a fast and nimble browser until version 2. After that Firefox has started to feel more bloated. Though 3.5 is faster than 3.0 in most areas, it has been out classed by safari and chrome.But until safari or chrome gets extensions it won’t be able to ween away the power users. Chrome and safari are at least 2 times as fast as Firefox. The tracemonkey JavaScript engine improved the JavaScript performance of Firefox drastically but it still falls short of that of chrome and safari.
This is what I had said in the last performance shootout.
Chrome and Safari are clearly the browsers to beat. What is surprising is that, with a relatively new JavaScript engine and a completely new architecture Google’s V8 has managed to beat the other JavaScript engines……we can definitely assume that Google is in it for the long term and has a strong interest and stake in making Chrome faster. So I’m making a bold prediction that in a year’s time, V8 will leave other JS engines in the dust
I still stand by that prediction. Though every browser improved upon their scores chrome has improved the most.
Browser Versions Used are Firefox 3.5, Safari 4.0.1 and Chrome Nightly Revision 19587. Sunspider is the benchmark suite of webkit (Safari) and V8 is the benchmark suite created by Google Chrome developers.
Performance Data
| Sunspider | V8 | |
|---|---|---|
| Chrome | 759.0 | 2698 |
| Firefox | 1527.4 | 259 |
| Safari | 863.6 | 1865 |
Get Updates via Email
About The Author
Manu is a Ruby on Rails developer,blogger, amateur entrepreneur, Mac and Linux enthusiast and a technophile. Read more ››
Get Updates via RSS feed
Popular Posts
- Download Native Google Chrome For Mac OS X
- Download Updated Google Chrome for Mac
- Google DNS vs OpenDNS: Google Rocks for International Users
- iPhone 3G Vs Nokia E71 Review
- Fix iPhone OS 3.0 Battery Charging Problems
opera？
I think my browser bogs down when it has 2-3 tabs open.
Can you do a test where you open multiple tabs and then run the performance benchmarks? Do the browsers perform the same under these conditions?
As mentioned in the post, opera was excluded because it has very poor
javascript performance. You can see my first javascript performance
shootout post (linked from the article) to see opera's scores.
2009/7/1 Disqus <>:
Which browser are you using ? 2-3 tabs is usually a very light load as
long as the sites you load are not heavy. I routines have 10-15 tabs
open and I see a performance drop only some times.
2009/7/1 Disqus <>:
I use all four browsers Chrome, Safari, IE and Firefox. Chrome is the best for me but Safari also works fast. Chrome has some good applications and book marking facilities. Bookmarking in Safari is also easy. New version of IE is worse for me. I rather use old version of IE for some purposes.
I tested it out on Safari on Mac and you are correct. The scores are
half of what it is without any other tabs open. I initially thought it
was because of flash and loaded only twitter, facebook, slashdot but
the effect was same. Browser was not bogged down but the scores came
down.
I did the test on chrome and chrome scores were the same as it was
without any tabs. This is due to the multi-process architecture of
chrome.
Shouldn't you also be testing against the WebKit nightly version of Safari? That way you could track Squirrelfish's progress against v8.
The reason I don't test webkit is that it is not clear what
improvements in webkit land up in Safari. Unlike chrome or safari
where a release is just a tagged nightly, wbkit nightlies never
directly become Safari.
I just did the same tests to compare Safari 4.0.1 with Webkit (r45463, although right now the latest one is r45641). Interestingly, Safari was MUCH faster than Webkit (and a bit faster than you report; would have been useful to know the computer in which you ran the tests). But in my case Safari was just opened, expressly for those tests, while Webkit has been running for weeks and has currently more than 40 tabs opened, so maybe that counts (although I didn´t expect that to be the case).
I'll repeat the tests (with freshly launched browsers) when I update to Safari 4.0.2 and the latest Webkit. (which I do when Webkit crashes, which usually takes weeks…). At that moment I will also try Chrome.
My exact results now were: (in a 2007 Macbook Core2Duo 2.16MHz , 3GB RAM, lots of background programs but about 10% CPU load)
Safari 4.0.1: V8: 2049 Sunspider: 696
WebKit r45463: V8: 1266 Sunspider: 4244
Sorry for hijacking your tests, I hoped this would be interesting.
Ran my tests on a macbook white 2Ghz Core 2 Duo with 2GB RAM
Actually having the browser open any other tab drastically reduces the
test performance in safari/webkit (I guess in firefox also). Chrome
due to its multi process architecture will give consistent results
even when there are other tabs.
Finally I updated all my browsers and repeated the tests. Yeah, Chrome wins…
Interestingly, Stainless (another Webkit-based browser which implements a separation between tabs similar to what Chrome does) is somewhat faster than both Webkit and Safari (though slower than Chrome), and not only does not get bogged down by more tabs, but seems to actually get a bit faster with the second or third tab. Chrome does the same. Go figure.
Details at my blog…
If you want to choose a car for your browser, which Car would it be? See here:
http://vamazing.com/index.php/comparing-cars-wi...
If you want to choose a car for your browser, which Car would it be? See here:
http://vamazing.com/index.php/comparing-cars-wi...
Bro, you need to tweak your firefox settings to reduce page rendering from 1 sec to 0 seconds to make 'at least seem' more snappy. Also, do not reduce memory when minimized, do not download favorite icons and disable blinking elements. I use a program called Auslogics BoostSpeed for firefox optimization but i'm sure you can make these custom tweeks on your own. Firefox runs like lightening once the appropriate tweaks are made. Chrome is still the fastest but only when surfing a singular page. Doesn't work so well when you have multiple tabs open and tends to freeze up after installing Windows 7(well worth the upgrade and a far superior operating system to Vista and anything Apple has up there sleeve. But then i've always felt Apple's OS was garbage. A Windows machine, when properly configured will outperform an Apple any day, simply b/c of all the registry tweaks that can be made in Windows). To be honest, IE 8 combined with new Windows 7 is probably the fastest browser on my computer. You have to have windows 7, make some tweaks for ex: allow more simultaneous downloads etc…as well as disable all the visual and animation effects and auto updates that slow IE down. I've hated IE but it is the only browser that can handle sony's new flash website without producing an error. I still don't use it b/c it's just ugly; microsoft's main pitfall; their gear isn't aesthetically pleasing. Safari is probably my favorite in terms of looks and ease of use but for whatever reason, it runs slow on my computer after a few tabs are open. Opera is close enough to Firefox but given the number of addons for firefox that are actually useful, I can't see myself switching to Opera 10. Have you given firefox 3.6beta 3 or 3.7alpah 1 a try? They are significantly speedier. But yeah, take home point with all these browsers is that it's all about the tweaks you make based on the kind of system your running and the internet connection you have. You can also optimize your internet settings too but again, I would use a program like Auslogics to do this. Massive speed gains. Hope some of this ramble helps.–Alex
I have used firefox with and without tweaked settings. I have no complaints
about the page rendering time. What I face are
1. It tends to chew up CPU once in a while forcing you to experiment close
some tabs
2. ditto but with memory
3. Complex javascript apps run much more slowly than chrome. No amount of
tweaking can help here. Try google wave in chrome and firefox and feel the
difference.
Will see how well the 3.6 fares.
I am working on a massive project that can only utilize jQuery library since it's presented over file protocol. I'm not able to discern much between Chrome 3's and Safari 4's performance, even on Windows (Firefox 3.5.5 lags even on when calling a small offset function with the cursor). However, I do notice that Safari's visual appearance, dithering, and animations are noticeably smoother than Chrome. I would assume, since both are built on Webkit, that this is the main cause for Chrome's faster performance?
Sorry, I am a *little* late. Too much time spent working. I just realized Google is using their own JavaScript engine, V8.
Firefox “outclassed” ? Not if you like standards.
Leave a comment