Subscribe via e-mail: 
Last year, when Firefox 3.5 was released and the whole net community was amazed by the performance gains of the Open Source browser, I was only amused by the childish obsession many IT pros have with browser speed. Now, as the Internet Explorer 9 developer preview is available, touting its fabulous hardware acceleration, I am beginning to wonder if the whole net is just nuts or if I should see a doctor because something is profoundly wrong with my sense of time. I am really worried. So please help me diagnose this phenomenon!
Of course, the first thing I tried after I installed IE9 was the speed demos that Microsoft kindly placed on the default home page. Even better is the Debug menu that allows you to switch between different Internet Explorer versions. It is common practice in browser benchmarks to test browser capabilities that play no role in real cyber world, so I was not surprised to see some browser logos flying and rotating on my screen. It was really a nice idea to use the logos of competitors in this demo, which shows how relaxed the IE developers are even though they are bashed all day by all those relentless IE haters and furious Firefox getters out there.
For the sake of those beaten IE programmers I hoped to see some significant speed differences in those funny performance demos. I switched and switched but I could not make out any speed difference no matter how hard I tried. I checked the load on my computer, but my CPU was only bored to death by these tests. So I decided to meditate a little in the hope that this would recalibrate my senses. And surprise, surprise—I then could really perceive a speed difference. I could swear that Internet Explorer 5 was faster than all the other browser versions, including IE9. Really! I could really feel it!
But then, I realized, this is not what those tests were all about. This was the point when I really started to worry about my perceptive faculty. I cleaned my glasses and looked in the mirror to see if my eyes looked unusual. I also tested my knee jerk reflexes and tried to point at the tip of my nose with closed eyes. Nothing. All performance tests ran through without any complaints and with the expected speed.
Well, I am really helpless now. Perhaps you could help me? Can you show me a web application where I can see and feel all these wonderful speed differences that everyone perceives? Please, though, no more benchmarks. I have no doubt that computers can exactly measure how long it takes to process a certain program. Show me a web application that you use every day where even your half-blind grandmother could see the difference between the sluggish Internet Explorer and the speedy Chrome browser. What I need now is a lift. I want to be a real member of the browser performance community. I also want to say wooooow, I can really feeeeel the speed! Please help me!!!
Erm, Michael, the IE 5/6/7/8 settings are for the render mode used (Emulating quirks of I5/6/7/8 in order to not break web sites). They don’t affect the speed in which these HTML5 demos are played.
You’ll need to compare the speed to Chrome, IE8, Opera and Firefox 3.6. I think Opera is the only one that can keep up with IE9 in this regard.
As for the general speed difference – i have no idea why you don’t notice it. Maybe you have another bottleneck besides the browser?
Do you have a source for this? It was also my first thought that the different rendering engines are only for debugging, but then I realized that I can’t move the rotating logos when IE5 is running.
But it doesn’t matter anyway. I never ever experienced any differences in browser speed. I would be very thankful if show me just one real world application where Chrome performs faster than IE. I don’t think that bottlenecks are the reason because it is always the same no matter what computer I use.
What I need is something that I could just measure with a stop watch, say a Google Maps map that builds up one second earlier or something like this. If you just tell me that Google Maps “feels” faster in Chrome, it wouldn’t really solve my problem.
Google Reader is faster in Chrome than in Firefox (Windows 7 64). I’ve never measured the exact speed, though. I guess it’s a heavy load for the browser: I have a hundred feeds with sometimes a few thousands unread items.
I use Google Reader every day and I can assure you that I read a ton of feeds. So I tried Google Reader this morning in Firefox, Chrome, and IE (Win 7 x64). Funny thing is, I couldn’t see any difference between IE and Firefox but Chrome hanged a couple of times. Shall I conclude now that Chrome is a sluggish browser? Certainly not. Obviously Google’s servers or perhaps my network connection hanged for a moment when I tried Chrome.
By the way, Google Reader is a very fast RSS Reader simply because it only loads as much titles as necessary to fill the screen. I couldn’t even measure the time it takes to load a page so fast is this thing. So perhaps Google Reader is not a good a choice if you want to prove that there is a real world speed difference between browsers.
Lukas, by the way, I just saw that the status bar displays the rendering engine and it also changes when you run the speed demos. If this isn’t proof enough that the whole talk about browser speed is only a joke then I don’t know what.
You have to look a little bit further than today’s Web pages. Vendors position their browsers more and more as front ends for applications, not just as document viewers. If you remember the announcement of Chrome, better execution for cloud apps was the main reason for Google to launch its own browser.
May be you don’t feel much of a difference when running a RSS reader in various browsers. But that will change with Web based spreadsheets, presentation software or picture editors. They do a lot of local processing and that is where you need a fast Javascript engine.
Wolfgang, I am quite aware that this is all about web applications. My point is that there currently are no web apps where browser speed matters. It remains to be seen if this will change in the future.
Considering that most PCs are overpowered when it comes to CPU speed and that web apps are comparably lightweight, it is not very likely that browser speed will matter in the near future. I mean, computation speed doesn’t even matter for fat PC apps, so why should this be different for those tiny web apps where quite a bit of the calculations can be done in the cloud?
Just installed IE9 beta. Two pages are MUCH faster: Google Maps (in scenarios with long routes, lots of markers, or interactively editing route), and Google Docs (spreadsheet).
Anyone see irony in the apps that are faster?
IE9 looks like it’s going to be a very promising browser. About time though