Well, MS finally did it (browser wars)

[Cogman]

Microsoft has finally come up with a browser test that Chrome can't hack http://ie.microsoft.com/testdrive/Performance/PsychedelicBrowsing/Default.html This brings chrome to a crawl (as well as IE8 and lower.. but that is beside the point). Lets see how long IE9 holds the browser performance crown.

 

[lxskllr]

FF 3.7=3RPM
FF 4b3=4RPM
Opera 10.6=5RPM
Chrome5=4RPM
IE8 Wouldn't run, errors on page

 

[atreader]

6.0.490.0 canary build (with 12 tabs open for 24 hrs): 9RPM
20100812 Minefield/4.0b4pre (fresh start, 1 tab) : 17RPM

at resolution/browser size 1600x800

Vista.

 

[ChAoTiCpInOy]

What speed is it supposed to be?

 

[Cogman]

What speed is it supposed to be?

I get 1789 rpms With IE9...

 

[ChAoTiCpInOy]

I get 1789 rpms With IE9...

Goddamn, seems rigged.

 

[llee]

Lol at the disclaimer at the bottom:

If you experience any symptoms, immediately consult a doctor and switch to a slow browser.

 

[Cogman]

Goddamn, seems rigged.

Not rigged, IE9 is using hardware acceleration, IE the video card, to beat the competition. No doubt they have a directx layer doing the rendering.

 

[VinylxScratches]

Still don't trust no IE as my main browser.

 

[Cogman]

Still don't trust no IE as my main browser.

neither do I. Like I said in the OP, it is a specific test made by microsoft. Of course it shines a light on chromes weakness. (You aren't likely to see webpages with extreme canvas rendering action)

However, my bet is that google will probably get to work on doing hardware acceleration under the hood (I can't say the same for the mozilla foundation, they have been somewhat slow to adapt)

 

[atreader]

Still don't trust no IE as my main browser.

Once in a while I check my almost dormant hotmail (windows live) account. Tried it yesterday and had to try 3 different browser to finally successfully forward an email. In one browser it just showed mobile version of hotmail. From my day to day web surfing I think only hotmail/windows-live site gives issues in non IE browsers (Opera 10.6, minefield, firefox 6.8, chromium), most other sites that I visit works just fine in these browsers.

A top company like Microsoft with a hugely popular site (hotmail) should be able to make a web compliant (website that works on common browsers) website.
/rantend

 

[Cogman]

Once in a while I check my almost dormant hotmail (windows live) account. Tried it yesterday and had to try 3 different browser to finally successfully forward an email. In one browser it just showed mobile version of hotmail. From my day to day web surfing I think only hotmail/windows-live site gives issues in non IE browsers (Opera 10.6, minefield, firefox 6.8, chromium), most other sites that I visit works just fine in these browsers.

A top company like Microsoft with a hugely popular site (hotmail) should be able to make a web compliant (website that works on common browsers) website.
/rantend

I use chrome pretty much exclusively and have never had a problem with hotmail, or any microsoft website (short of maybe windows update, but that really doesn't count)

 

[atreader]

I use privoxy along with those browsers, maybe that could be the reason. Also browsers are set to 'Private Mode'. I don't have problem accessing gmail or yahoo mail.

 

[bunnyfubbles]

 

[Cogman]

Yes yes, decrease the resolution, increase the rotation speed, not a perfect test, I get that.

 

[bunnyfubbles]

Yes yes, decrease the resolution, increase the rotation speed, not a perfect test, I get that.

considering the test actually notes the browser window size, I'd think they'd have accounted for it, but I guess they were just super lazy. Considering shrinking the window size from ~1680x1050 to 500x430 goes from single digits to one thousand+, I'd wager the difference in GPU acceleration would be just as huge.

Thus while the goal may be to get >1000RPM @ full resolution, if you're getting "only" 1000RPM @ 500x430, the browser is still slow compared to what it could be with GPU acceleration.

 

[Cogman]

considering the test actually notes the browser window size, I'd think they'd have accounted for it, but I guess they were just super lazy. Considering shrinking the window size from ~1680x1050 to 500x430 goes from single digits to one thousand+, I'd wager the difference in GPU acceleration would be just as huge.

Thus while the goal may be to get >1000RPM @ full resolution, if you're getting "only" 1000RPM @ 500x430, the browser is still slow compared to what it could be with GPU acceleration.

surprisingly it isn't. Ran the same test with IE9 at a reduced resolution, on this computer it stayed at about 1800rpm

 

[KeypoX]

wow i only got 13...

ahh nevermind you just made your window smaller