Tumblr 101: Tips And Tricks To Get Started
Sebastian Thrun: Google's Driverless Car
Hype Check: Safari 4 can't beat Google Chrome
Perhaps they're referring to a different world. On the planet Earth, however, Safari 4 still can't surpass Google Chrome on the Sunspider, V8, or Dromaeo browser benchmarks.
If you'd like to see how it stacks up, take a look at my comparatives from Dromaeo - Safari 4 is the column labeled Webkit 528.16. Other browsers used were Chrome 2.0.160.0 (labeled Webkit 530), Chrome 1.0.154.48 (labeled Webkit 525), Firefox 3.0.6, Firefox 3.1 (labeled 60780) with Tracemonkey enabled, Opera 9.63, and Opera 10. It's no longer news how slow Internet Explorer (both 7 and 8) are at processing Javascript, so I haven't included its results here. Tests were run on an AMD quad 9550 with 4gb of memory running Windows Vista Ultimate x64.s?
Ugly graphs after the break...
Safari 4 does perform admirably when tested on Dromaeo, placing first in ten tests. Firefox 3.1 claims seven top marks and comes close on two more. Based on overall numbers, however, Chrome 2 still wins the Javascript speed title.
Wait, didn't CNet just release "shocking results" that show Safari 4 coming in first? Sure they did. And Webware ran a post showing it coming in second in their battery of tests. I'm going to have to side with Webware on this one.
At this point, though, what do the numbers mean? Firefox 3.1 puts up a score three times the next best in the "bits in byte" test. Chrome 2's 1057.765 in recursive number calculation is about 25 times what Firefox can manage. Safari's Fannkuch result is 33% better than that of Chrome. Each browser seems to be very strong in certain areas and weak in others, so do the numbers really matter that much?
No.
They're interesting to look at, but in actual use Chrome 1 and 2, Safari 4, Firefox 3.1, and even Opera 10 "feel" very similar in terms of performance to me. Ultimately, it boils down to using the browsers yourself to determine which one is best suited to your browsing habits.
Ignore the hype machine. Who cares what browser puts up the best test results if it doesn't pass muster on your favorite sites.
Comments
17
Subscribe to commentsstuartmFeb 25th 2009 7:48AM
Safari beats Chrome in the Sunspider test on my EeePC 1000h.
KarlWFeb 25th 2009 9:18AM
Yes, performance metrics vary with the computer you're running. Especially with these new JS engines that create native code.
Perhaps Apple's result was based on testing the OSX version against the windows Chrome. Not ideal, but until chrome gets a native OSX version, one of them is going to be running with compatibility libraries (I don't rate the performance of Apple's Cocoa -> Windows library)
maf teechrFeb 25th 2009 2:28PM
From Safari's download page:
"Testing conducted by Apple in February 2009 on a 2.8GHz Intel Core 2 Duo-based iMac 24-inch system using Windows Vista SP1 and configured with 2GB of RAM and an ATI Radeon HD 2600 PRO with 256MB of VRAM."
They used an Apple computer but on a Windows OS.
KeegdnaBFeb 25th 2009 8:30AM
I don't care if it is faster (which, apparently it isn't)
Safari is still Safari and Safari wil ALWAYS suck. I dislike it about as much as IE.
Firefox, Chrome, and Opera FTW
synerageFeb 25th 2009 10:06AM
Uh, why? Safari's a very well-made browser and they've fixed up the Windows version a lot with Safari 4.
Lee MathewsFeb 25th 2009 8:34AM
Agreed. That's why I suggest testing the browsers yourself.
If you live in an area with a lot of gravel roads, a Ferrari isn't a great choice for you, even if it is speedy.
JohnnyFeb 25th 2009 9:01AM
Ummm.... maybe its because you're only testing the Javascript engines? I think it goes without saying but y'know the internet doesn't just run off javascript, the performance of the *actual* rendering engines sill has to be taken into account and can have a noticeable impact.
Take Opera 10.0 for example. Opera haven't got their next-gen javascript engine out of the gate yet (in any form), and yet Opera 10.0 still manages to load quite a few pages faster than Chrome, thanks to its rendering engine.
MrViklundFeb 25th 2009 9:32AM
Can't? Can't beat? What a FAN BOY article...
Lee MathewsFeb 25th 2009 9:31AM
It might be worth noting that I use Firefox, not Chrome. Chrome doesn't handle our CMS system properly.
If anything, I'm advocating per-user experimentation, not one particular browser.
mr.macFeb 25th 2009 10:27AM
Karl W is right, Safari on OSX runs much faster and with less bugs than the windows version.
I would agree that Safari on Windows sucks, it crashes alot and uses to much memory, but on OSX it is as slick as Oil.
AdamFeb 25th 2009 10:47AM
Does it even matter?
Javascript engines have become so fast, the performance differences are incredibly difficult to notice during normal usage. In fact, to see the difference, they have to put it through these benchmarks to stress it.
I can see a difference between IE 6 and Chrome when using GMail, but I can't see a difference between Safari and Chrome for the same test.
michas_piFeb 25th 2009 11:35AM
I like Safari 4 beta a lot but until it supports the extensions I use with Firefox, I'm staying away.
Wide SpectrumFeb 25th 2009 3:05PM
I could have sworn this was a beta version! As in unfinished. Oh well!!!
Lee MathewsFeb 25th 2009 3:06PM
It is, but so are two of the others I compared it to (Chrome 2, Firefox 3.1).
jimFeb 25th 2009 4:20PM
Google Chrome is spyware, Safari wins.
Well actually Firefox wins since Safari and Chrome both suck. :P
mxxconFeb 25th 2009 4:52PM
here are results on my computer
safari4 beta vs latest dev Chrome
http://dromaeo.com/?id=60813,60818
VMMasterMar 12th 2009 4:54PM
Yeah but the real question lies in when the EU is going to make Safari un-installable in OSX