Web Browser Grand Prix: Chrome 25, Firefox 19, And IE10

[mayankleoboy1]

^ +1
And since the Opera as we know it is dead (to be replaced by ChrOpera), stacking odds in favour for it as a martyrs tribute would have been OK, if only for sentimental reasons.

 

[Pherule]

Chrome is great, in theory. Except it downloads and installs updates without your permission and without in-browser customization options for stopping this behavior.

 

[Guest]

i find that chrome seems to be smoothest browser for me, what do yall think?

 

[jeffredo]

I gauge the best browser on how well it can handle ancient hardware. Recently re-assembled an old Athlon XP PC from a box of upgrade leftovers that I've been collecting for years. Chrome is the only one that out of the four that provides a reasonable satisfying browsing experience not too much different than my modern gaming rigs. On the other three the tired of XP 3200+ struggles with busy webpages.

 

[tipoo]

Sad to see Opera flounder so much, but I guess that's why they're going to be switching to Webkit rather than their proprietary engine. The beta Opera with Webkit seems pretty good on Android, I'll be excited to see the desktop one.

However as for Firefox earning second place, I just don't see it in personal use, it may come in second in benchmarks but in perceived speed it always seems the worst to me.

 

[adamovera]

[citation][nom]slomo4sho[/nom]I currently use Firefox, Chrome, and IE 10. They each have their own advantages. Thanks for the write up. When can be expect the Android browser round up?[/citation]
The ran towards the end of last year. I'm basically waiting on new developments to do another.

 

[adamovera]

[citation][nom]xintox[/nom]Really? I thought you were better than that at Toms.The composite hardware acceleration scores is most likely the main reason why IE9/10 is so far behind Firefox and Chrome on performane. Yet, from what I can tell, this composite score is heavily influenced by the WebGL scores, which is exclusive to Chrome and Firefox.In that respect, MS has at some point stated that they do not even wish to support WebGL, as it represents a significant security risk, as it gives the browser close access to the computer hardware.Long story short, your methods of calculating performance scores heavily favors Chrome and Firefox as they are the only ones to implement support for WebGL.Additionally, I wish you would make it more clear how you arrive at your composite scores and of course the final Performance Index. How do you add numbers that are so varied in nature, without some method of normalizing the numbers?[/citation]
The HWA scores do favor Chrome and Firefox because they are the only browsers to have WebGL enabled by default. IE isn't going to get a pass by willingly not adding functionality that all of its competitors will eventually have (Opera and Safari have WebGL implementations as well, just not enabled by default in the stable release like Chrome and FF).
The composites are usually either simple averages or geometric means. Averages work when all the tests in question use the same scale. Geometric means come into play whenever we have differing scales in the same category of testing.

 

[adamovera]

[citation][nom]srap[/nom]@adamoveraI heard that Futuremark Peacekeeper is unreliable, it used to miscalculate it's own benchmarks and it's still a black box so it may still be buggy. Do you know anything specific about it?[/citation]
New to me, Peacekeeper 2.0 dropped the beta tag awhile ago. As far as the black box goes, it's hard to tell exactly what that test is measuring, especially if reading the marketing-speak on the main page. Though the FAQ reveals that it's simply another JS test. All those extra tests that it runs aren't factored into the final performance score.

 

[adamovera]

[citation][nom]abbadon_34[/nom]Nice to see IE9 AND IE10, as well as Win 7 AND Win8. But I am curious how the final scores were calculated, seeing how WebGL was only supported on 2 browsers, as well as how the other incompatible tests were scored. A 0 vs not including can make a good deal of difference.[/citation]
The other three browsers were given a score of 1 in the WebGL tests so as not to completely wreck any positive performance they exhibit in native HWA.

 

[adamovera]

[citation][nom]belardo[/nom]"Before Opera becomes yet another Webkit browser and we're down to the top three," What does that mean? That its taken off the list because it'll be a webkit? Chrome is already a webkit.[/citation]
It's possible, if the initial version of the new Opera is clearly just another spin of Chromium with a twist (e.g. Maxthon, Rockmelt, Comodo, etc.) than I see no reason to keep running it - experience has shown that such browsers can only hope to match Chrome. On the other hand, if they really punch it up in some way (like Dolphin with its new Jetpack HTML5 engine), we'll most definitely keep it. Can't say for sure one way or another at this point in time, but from what little we know, it is very possible that Opera will become another mini-Chrome with an untestable gimmick of some kind, in which case there's no point in keeping it in the running.

 

[adamovera]

[citation][nom]ET3D[/nom]I'm impressed by the difference between Windows 7 and 8.Also, FF is still the winner for me. The memory footprint with a lot of tabs open is pretty important for me. My wife's PC (3.5GB, Windows 7 32-bit) regularly bogs down when she opens a lot of tabs in IE. Have all the javascript and HTML 5 performance you want, when RAM runs out the entire computer slows down to a crawl.[/citation]
I can say that the last few versions of Firefox have felt faster in daily usage than most of the 4.0+ series. Although I switched to Chrome as my primary browser around version 3, the TH CMS keeps me using Firefox regularly, and I can see a difference in the startup and responsiveness lately, especially on Windows 8 (my laptop ran 7 and 8 back-and-forth for about six months).
As far as total memory usage goes: these figures are not absolute, meaning they will change depending on available system resources. It doesn't actually take Chrome 1.5 GB to display 40 tabs, it just uses that much on this ample test system. It only uses about with just 786 MB of total system RAM - to load the exact same pages! This is why we make the final efficiency determination based on how close a browser comes to its single-tab usage after closing the additional 39 tabs. Total usage is completely dependent on available system resources, and this is especially true for Chrome.

 

[adamovera]

[citation][nom]shahrooz[/nom]I ran some of the tests on maxthon 4 and I got some good results it has a nice look too I'm going to test it for a while, anyone else has and experience with it?[/citation]
We were actually pretty serious about including Maxthon in a WBGP when it looked like IE10 for Win7 was delayed. It's performance was not quite at the same level as Chrome, or even Firefox once the newer technologies entered the picture. Unfortunately, it had trouble running Kaizoumark (which we had to table this time anyway for IE9) and it would not complete BrowsingBench at all due to a bug which caused it to start downloading portions of the test pages!
Maxthon is a very slick browser, I love the UI, and the iOS version is best thing there is on that platform other than Safari. We'll probably need to do a separate B-list WBGP with a modified test suite that can handle all the second-string browser's quirks - a ton of benchmarks only make sure to function on the Top 5.

 

[adamovera]

[citation][nom]Philippe Leblanc[/nom]It would have been nice to see some results with opera's HWA turned on. I can understand why you wouldn't count those results towards the final score since you are looking a stock browser configs. But the numbers make opera look so bad because there is no HWA by default. However, going into opera:config and turning HWA and WebGL on and run benchmarks just for reference would help opera's cause. They've made a lot of progress in that department and showing that would allow people to see how fast opera's current HWA implementation is. Just because the developers didn't deem it ready for primetime doesn't mean they shouldn't get a little credit for their work up to now. Just my 2 cents.[/citation]
[citation][nom]mayankleoboy1[/nom]^ +1And since the Opera as we know it is dead (to be replaced by ChrOpera), stacking odds in favour for it as a martyrs tribute would have been OK, if only for sentimental reasons.[/citation]
I'm OK with that, I'll turn it on in Opera and Safari next time to see where they are with it, but not in the final scores.

 

[adamovera]

[citation][nom]jeffredo[/nom]I gauge the best browser on how well it can handle ancient hardware. Recently re-assembled an old Athlon XP PC from a box of upgrade leftovers that I've been collecting for years. Chrome is the only one that out of the four that provides a reasonable satisfying browsing experience not too much different than my modern gaming rigs. On the other three the tired of XP 3200+ struggles with busy webpages.[/citation]
Opera was no good? With more than a few pages, , though Chrome is also solid.

 

[adamovera]

[citation][nom]adamovera[/nom]The ran towards the end of last year. I'm basically waiting on new developments to do another.[/citation]
The "first Android WBGP" ran towards the... I had a link but I guess that it's too much to ask that it gets posted in here:
http://www.tomshardware.com/reviews/windows-xp-web-browser-performance,3167-13.html

 

[adamovera]

[citation][nom]adamovera[/nom]Opera was no good? With more than a few pages, , though Chrome is also solid.[/citation]
Not even sure what I said in this link, sorry:
http://www.tomshardware.com/reviews/windows-xp-web-browser-performance,3167-13.html

 

[blazorthon]

[citation][nom]srap[/nom]Because if it switches to webkit, all of it's test results will be identical to Chrome's, or it will hardly differ.Firefox 15: HueyFix, details here.@alidan: FF17 ESR would be more safe than FF10, and even better with memory.Edit: I hate how it keeps deleting the links.[/citation]

Safari is also Webkit. It doesn't work like Chrome. Opera going to Webkit doesn't mean Opera being a clone of Chrome.

 

[mayankleoboy1]

[citation][nom]blazorthon[/nom]Safari is also Webkit. It doesn't work like Chrome. Opera going to Webkit doesn't mean Opera being a clone of Chrome.[/citation]

+1
Actually, if you get a little deeper into Safari and Chrome development, you will find that "Webkit" is not completely open source! Some portions of it are held as closed source by BOTH Apple (no surprises, huh?) and Google (surprise!) . So the Webkit of Safari != Webkit of Chrome. Its more like there is an Apple Webkit, and there is a Google Webkit.

Really, Firefox is the only True Open Source Browser.

 

[adamovera]

[citation][nom]blazorthon[/nom]Safari is also Webkit. It doesn't work like Chrome. Opera going to Webkit doesn't mean Opera being a clone of Chrome.[/citation]
[citation][nom]mayankleoboy1[/nom]+1Actually, if you get a little deeper into Safari and Chrome development, you will find that "Webkit" is not completely open source! Some portions of it are held as closed source by BOTH Apple (no surprises, huh?) and Google (surprise!) . So the Webkit of Safari != Webkit of Chrome. Its more like there is an Apple Webkit, and there is a Google Webkit.Really, Firefox is the only True Open Source Browser.[/citation]
If the initial version of the new Opera is clearly just another spin of Chromium with a twist (e.g. Maxthon, Rockmelt, Comodo, etc.) than I see no reason to keep running it - experience has shown that such browsers can only hope to match Chrome. On the other hand, if they really punch it up in some way (like Dolphin with its new Jetpack HTML5 engine), we'll most definitely keep it. Can't say for sure one way or another at this point in time, but from what little we know, it is very possible that Opera will become another mini-Chrome with an untestable gimmick of some kind, in which case there's no point in keeping it in the running.
Opera did say Chromium specifically, not simply WebKit, and Chromium means the WebKit rendering engine and the V8 JavaScript engine. Safari uses WebKit with a different JS engine, and AFAIK, it uses a newer version of the rendering engine as well.

 

[srap]

Edit: disregard this, adamovera hijacked me

 

[loftie]

You know Opera has Hardware Acceleration disabled by default. I was expecting you guys to test with it on and off to show the differences.

 

[mayankleoboy1]

[citation][nom]adamovera[/nom]If the initial version of the new Opera is clearly just another spin of Chromium with a twist (e.g. Maxthon, Rockmelt, Comodo, etc.) than I see no reason to keep running it - experience has shown that such browsers can only hope to match Chrome. On the other hand, if they really punch it up in some way (like Dolphin with its new Jetpack HTML5 engine), we'll most definitely keep it. Can't say for sure one way or another at this point in time, but from what little we know, it is very possible that Opera will become another mini-Chrome with an untestable gimmick of some kind, in which case there's no point in keeping it in the running.Opera did say Chromium specifically, not simply WebKit, and Chromium means the WebKit rendering engine and the V8 JavaScript engine. Safari uses WebKit with a different JS engine, and AFAIK, it uses a newer version of the rendering engine as well.[/citation]

You probably wont, but can you please update these benchmarks (but not the final result) with HWA enabled in Opera ? The next WBGP will be quiet far in the future, and by then Opera 12.14 will be ancient history.

 

[cirslevin]

you measure the memory efficiency, but not memory usage? why? do user's computer got bogged down by lacking of memory? or by lack of memory efficiency?

And you booted all open source js benchmark, in favor of all black boxed benchmark? why?

Your chrome bias is off the chart now.

 

[fkr]

but im going to help some people having memory issues in opera

my RAM is 2GB and heres my opera:memdebug configuration

http://i.imgur.com/uoCsi.jpg

some opera:config setup that i did

-Disable "Cache Other" under disk cache tab(this means opera wont cache "OTHER" content like flash videos,mp4 and some multimedia content..downside is that in opera:cache you cant check those videos and audio to search for it.
-Disable Thumbnails
-Disable bittorent
-Disable unite,link
-Disable UI animations
-Disable Special Effects

and if you want ultimate reduced memory usage..
-Disable "Cache Docs and Cache Figs" both under Cache and Disk Cache tab in opera:config(NOT RECOMMENDED to me since it will make opera very slow and not cache'ing content in disk cache and RAM)

as im typing this i have 11 tabs open(2 YT videos the rest is document and images sites) and only using 232 MB of RAM right now.

hope kinda help other people.

 

[mayankleoboy1]

^ So what do you do with the 600MB spare RAM ? Put it in a showcase for everyone to see and admire ?