HTML 5 Speed Test: IE9 vs. Chrome 6

The Chrome 6 framerate swings wildly in that video. 1000 fish in the IE9 preview 3 does 23-24fps at 1920x1200 on my system. The JS performance in IE9 preview 3 is 10-15% faster than the current FF and not horribly slower than Chrome 5 anymore (but still slower). It's about time, MS. Now finish it up and release it.
 
I wonder how much running the two browsers simultaneously hurt IE9's performance once he got to 1,000 fish. I use Chrome every day at work and at home, and I might be switching over to IE9 if the performance difference stays this large.
 
So, it looks like Chrome is a turd?
Compared to browsers with hardware acceleration, while using the HTML5 canvas, relatively yes. I can't imagine that Chrome 6 final will lack hardware acceleration.

BTW, the captured video just looks jerky. The fish demo in IE9 is smoothly animated when you run it locally.
 
very interesting.....maybe ill give it a go and move away from chrome.


could it be microsoft coded the page to run better on IE9 than the competitors? (....its a microsoft "test drive" page if you look at it closely), no idea if this is possible just a thought
 
Chrome will catch up and possibly end up surpassing IE9's Canvas rendering performance. Of that I have no doubt.

I wonder how the new Safari fares, though. It's now hardware accelerated in Windows as well.
 
so if I appreciate the speed of HTML 5 > Java, does that mean I agree with Steve Jobs?

(I fink I dust puukt a wittle in my mouf ) :(
 
Chrome will catch up and possibly end up surpassing IE9's Canvas rendering performance. Of that I have no doubt.
I do have a doubt about that. :p IE9 only has to run on Windows, and can use whatever MS wants that is Windows exclusive without regard to backwards compatibility (i.e. Direct2D or other DX APIs, no XP support).

I just read that Chrome does have hardware acceleration in the version 6 development branch. Oops. Using OpenGL to work cross-platform does have a disadvantage.

Anyways, here's an updated 4 way shootout from the same guy: http://www.youtube.com/watch?v=OIZUdZdFzOo&feature=player_embedded

FF 3.7 and IE9 have about the same performance, and both happen to use Direct2D. :p Opera is about 1/2 the speed of those and Chrome comes in last.
 
Normally I roll my eyes when browser developers claim that they're the fastest browser when in these days it only means about 1/20th of a second, but THIS deserves recognition for speedy browsers.
 
So whats the point of this again? Rendering fish faster? I'll stick with Chrome and wait for Chrome 6 to come out. :)
 
I just read that Chrome does have hardware acceleration in the version 6 development branch. Oops. Using OpenGL to work cross-platform does have a disadvantage.
And it's still quite early. We don't really have a firm idea about what state Chromium is at this point when it comes to its hardware acceleration implementation. To suggest that OpenGL is a performance hindrance without having a solid idea about the state of the implementation itself is a little premature.
 
So whats the point of this again? Rendering fish faster? I'll stick with Chrome and wait for Chrome 6 to come out. :)

Lol, I know..

Hrmm lets use benchmark specifically designed for one browser against a browser that wasn't designed to run it and hasn't been fully developed yet.... how about some real world scenarios?
 
Lol, I know..

Hrmm lets use benchmark specifically designed for one browser against a browser that wasn't designed to run it and hasn't been fully developed yet.... how about some real world scenarios?

lol I know because IE9 is completely developed :rolleyes:
 
Hrmm lets use benchmark specifically designed for one browser against a browser that wasn't designed to run it and hasn't been fully developed yet.... how about some real world scenarios?
LOL at the ridiculous statement. :rolleyes:

The demo runs flawlessly on serveral HTML 5 capable browsers (FF 3.7, IE9, Chrome 6, Opera), using JS and the HTML 5 canvas. Safari wasn't tested, but it doesn't support HW acceration under Windows at all. So it's kind of pointless to include it since it will be a slide show with very few objects. If Apple decides to one day enable HW acceration for Windows, it will be interesting to compare it with the rest.
 
Shoot and me with out a fishing pole.

From what I can tell by this test, IE9 supports more little fishes than Chrome. But Chrome can make your small fishes grow large in a matter of seconds.
Hmmm, more fish or bigger fish. That is the question.
 
Safari wasn't tested, but it doesn't support HW acceration under Windows at all. So it's kind of pointless to include it since it will be a slide show with very few objects. If Apple decides to one day enable HW acceration for Windows, it will be interesting to compare it with the rest.
It's available now. Safari 5.
 
Safari 5 only supports HW acceleration in OS X. It does not support HW acceleration in Windows. Plz read more carefully.
Follow your own advice, my man:
Hardware Acceleration for Windows
Tap into the graphics processing power of your PC while browsing the web. Safari 5 adds hardware acceleration support for Windows, so rich media and interactive graphics execute smoothly and speedily in the browser.
http://www.apple.com/safari/whats-new.html
 
:rolleyes: Again, not for the HTML 5 canvas object.

The Flying Images demo results again surprised us. Apple specifically mentioned limited GPU acceleration as one of the features for the Windows version of Safari 5. However, Safari tied for dead last place on this test with Chrome—which hasn't done much if anything to leverage GPU acceleration—at a measly 3fps. Firefox was far faster at 12fps, and Opera was far faster yet again with 28fps. However, IE9 turned in a mind-blowing 64fps. Again, given Apple's promotion of newer Web standards for animation, the fact that this demo runs 6 times slower on Mac OS X on much more capable hardware has to be embarrassing.
http://arstechnica.com/apple/news/2...d-chrome-opera-still-have-javascript-edge.ars

Get it now? The water is right there.
 
Again, not for the HTML 5 canvas object.
I can find no such evidence to back up this claim and it's not specifically mentioned in the article you linked to. Its poor performance alone isn't indicative of an absence of hardware acceleration.
 
If you choose IE9, the good news is that its super fast at flash rendering. But the bad news is, you'll be using IE9.
 
What am I looking at and supposed to be impressed / depressed by? It looks like a video of fish, with no meaningful differences between them.
 
If you choose IE9, the good news is that its super fast at flash rendering. But the bad news is, you'll be using IE9.
This is not flash. It's rendering to the HTML 5 canvas using Javascript. IE9 preview 3 doesn't even support plugins yet.
 
What am I looking at and supposed to be impressed / depressed by? It looks like a video of fish, with no meaningful differences between them.
You are hopefully looking at the start of the end of the horribly unstable and security threat called Flash for scripted animation and games.
 
Nobody does hardware accelerated graphics as well as Microsoft, I'll give them that.
 
I'd love it if MS hits a home run with IE9.
I'd prefer a browser directly integrated with UAC, feature rich...

I'd like to be able to ditch Firefox (I won't use Chrome because of Google... And yet I still host my email there, go figure...) but Microsoft is making it too difficult.
 
Back
Top