IE9, FF4 Beta In Real-World Benchmark

Source: http://www.lucidchart.com/blog/2010/09/16/ie9-ff4-beta-in-real-world-benchmark/

 

Why do these comparisons come up from time to time? Web browsers' raw processing speed is not the only thing from a users point of view, it is the whole experience of using the browser. So the real comparison should be which browser package gives the most to the users, this should include everything from processing speed, customizability, user-friendliness, usable GUI, privacy and security, available third-party plugins etc.

 

So the summary of this benchmark is that GPU acceleration sucks..?

I don't see how that's even possible. You can run the browser yourself and see that GPU accelerated browsing is faster, forgive me for being skeptical.

 

Where do these things come from and why do so many people hunt them down and post them? It seems like it does absolutely nothing but add more for fanboys to point at and puff their chests with. Any minute now the Googleheads will show up with their "facts" here and the naysayers will argue how the test is flawed, and it'll just go on and on like always. Doesn't ANYONE else but me see that these "tests" rely on a boatload more than just the browser? There's always different hardware involved, any real test is going to be flawed due to internet congestion, network glitches, that sort of thing. It's VERY hard to get any REAL facts from these things.

 

I'm of the opinion that all these benchmarking "tests" are completely and utterly useless. Yes, they may be interesting from a purely technological standpoint, but say nothing of the usability and experience of the browser in everyday use.

I mean, really, A is 2ms slower than B in one discrete area of the browsing experience? So what. That tells me nothing, absolutely zilch, about how I will like the browser.

Use them for development as a barometer, but don't shove the stats in my face and say the browser is the best ever because your tabs open .01 ms faster than the other guy when the moon is full and I stand on my head.

Really....

 

Well does this test have some meaning? Is it trying to tell us that GPU acceleration doesn't speed up the browser? It doesn't seem to be trying to tell us that, but that's the impression it's giving by saying firefox 4 is inferior to 3. Maybe I missed it? Did you understand what the message was?

Telling me that a non-GPU accelerated browser is superior to a GPU accelerated one just seems plain illogical. Am I wrong? Why else would Mozilla, Google & Opera be rushing hot on Microsoft's heels to add GPU acceleration if it's this terrible?

 

Honestly I'm not sure it was anything more than just another opinionated advertisement for a browser, as most of them inevitably end up being. As far as your question, I would hope they weren't saying GPU acceleration is "useless", as that's just plain silly. Now, whether it is far above and beyond in performance compared to a non-GPU accelerated browser, well, I'm afraid that too is a test that won't mean much as it for SURE will be dependent upon hardware.

 

Yep, pretty much. Who cares really.....

 

One glaring omission in this test is information about the GPU chipset in the laptop. Motherboard graphics are notoriously weak and laptops even more so.

I did some casual tests (Fish IE tank and Psychedelic Browsing from the IE9 test drive site) with IE9 and FF4b6 on my laptop and desktop. On the laptop FF4b6 was much faster. On the desktop IE9 was faster, but only slightly. A meaningful test would have to include a cross section of GPU's.

 

at least some typical PC setups (low middle high of both laptop and desktops). On my Son's gaming desktop IE8 was fast (cold <2 secs, cached <1 sec), he did not mind. On his university laptop Chrome gives him the best feel. On my play PC with much moderate hardware IE8 loaded slowly while Opera and Chrome/Iron run really well and FF's first startup took ages.