Firefox 9: The Duke Of JavaScript Performance

By | September 7, 2011


Firefox 9: The Duke Of JavaScript PerformanceExciting times ahead.

Good news for all the Firefox users and bad for its competitors, the upcoming Firefox 9 release will offer a significant JavaScript performance increase.

According to various reports, the latest Firefox 9 nightly builds score up to 32% more in JS benchmarks when compared to Firefox 6 and it’s not even finished yet.

Firefox 9: The Duke Of JavaScript Performance

Firefox 9: The Duke Of JavaScript Performance

Firefox 9: The Duke Of JavaScript Performance

Will all the GUI improvements and memory consumption tweaks in Firefox 7 and 8 releases, it looks like Mozilla is going back to the game.

Via: CT

[Thanks to everyone who sent this]


About (Author Profile)


Vygantas is a former web designer whose projects are used by companies such as AMD, NVIDIA and departed Westood Studios. Being passionate about software, Vygantas began his journalism career back in 2007 when he founded FavBrowser.com. Having said that, he is also an adrenaline junkie who enjoys good books, fitness activities and Forex trading.

Comments (35)

Trackback URL | Comments RSS Feed

  1. Guest says:

    Why would you not include Opera and Safari benchmarks? Other than Sunspider Mozilla doesn’t beat Chrome, one out of 3 benchmarks does not a winner make. 

    My Opera 12.0 pre- alpha clocks Sunspider @ 220 http://simplest-image-hosting.net/png-0-image-20110907-133508 (did I read this right?)

    The question that needs to be asked is what score need to be obtained for the browser experience to be “good” because at some point the performance increase won’t really be noticeable (except to those hyping their scores)

    • Sarjoor says:

      You are reading the image correctly, but just one image is useless.  All that image tells you is there is a fast computer that completes Sunspider in 200ms, presumably Opera from the above poster.

      But Javascript benchmarks are only useful when testing the speed of different Javascript engines all running from the same computer.  Without being able to compare the Sunspider speeds of other JS engines/browsers on the same computer makes that image useless.

    • Anonymous says:

      Indeed, all this proves is that consumers are fucking morons who all easily fall for cherry picked graphs showing a particular browser in a good light whilst totally ignoring the bad stuff.

      I’m surprised how many “tech” sites will actually destroy their reputation with this crud PR…

      I guarentee Firefox won’t feel 30% quicker than before, infact I bet it feels slower, as clearly the focus on making something work fast for a benchmark does not translate very well into how people USE browsers on a daily basis.

      How many people actually look at WebGL fishtanks all day?  I rest my case….

      • apriorimeister says:

        I love how you never have any proof to back up any of your statements. You just spout a biased opinion and all of these other Opera fanatics just ‘like’ your comments. Anyways, speaking of bias – I honestly don’t see much point in including the V8 benchmark. My point being is that the benchmark looks for qualities that Chrome is specifically optimized for and so it is unlikely that we will ever see any browser come close to Chrome in it. Pretty obvious… I know.

        • Anonymous says:

          Well Mozilla claim every new Firefox release is faster than the last one, but the reality is that it’s just more bloated and sluggish.

          The real point here, is Apple(Sunspider), Mozilla(Kraken), Microsoft(Testdrive) and Google(V8 benchmark) have all created benchmarks to show off certain strengths of their browsers, Opera didn’t need to, as anyone smart enough to use Opera knows benchmarks means jackshit in the real world.

          • apriorimeister says:

            Firefox does get faster after pretty much every ‘major’ release, that’s an undeniable fact. I also disagree with you that benchmarks don’t mean anything, otherwise we frankly wouldn’t even have them. Even Opera developers at the release of 10.50 and 10.51 were spouting about how great the performance was and in turn they showed us benchmarks: http://my.opera.com/chooseopera/blog/2010/03/22/opera-10-51-released-making-the-fastest-browser-even-faster The only reason why we aren’t seeing more benchmarks conducted and shown to the public from Opera developers is because it sucks in practically every single one of them. Peacemaker being the exception in which Opera usually takes the second place (after Chrome) and sometimes even the first.

    • Anonymous says:

      Indeed, all this proves is that consumers are fucking morons who all easily fall for cherry picked graphs showing a particular browser in a good light whilst totally ignoring the bad stuff.

      I’m surprised how many “tech” sites will actually destroy their reputation with this crud PR…

      I guarentee Firefox won’t feel 30% quicker than before, infact I bet it feels slower, as clearly the focus on making something work fast for a benchmark does not translate very well into how people USE browsers on a daily basis.

      How many people actually look at WebGL fishtanks all day?  I rest my case….

    • Anonymous says:

      Simply, it’s not worth. Waste of time. The test should include only those worth being mentioned.

      ’nuff said

  2. Tiago Sá says:

    Won’t make the slightest of differences. Mark my words.

    Firefox’s marketshare will take a very long long while to start growing again. It’s all about marketing, and Mozilla hasn’t it.

    • Anonymous says:

      I agree here.  Firefox has a stable base, but there is only room for one alternative browser.

      Windows dolts use IE
      Mac dolts use Safari

      Savvy users choose from Opera, Chrome or Firefox, but only one can ultimately gain widespread adoption.  My money is on Chrome, without a doubt.

      Firefox will struggle to maintain above Safari levels, and ultimately fall below as MacBooks become more and more predominant.

      As mobile devices become the only devices people use to access the Web, Firefox will be in a marketshare battle with Opera and probably lose there as well (Opera is so deeply-rooted as an alternative mobile browser).

    • Anonymous says:

      Although, I get your point, you’ve to realise that products targeted around the world, like Firefox and Chrome, ain’t so easy to promote. Especially in times of recession. Comparing the browsers currently, only Google is able to promote Chrome as you mean. You understand, I guess, why so.Mozilla have mentioned many times that hasn’t the recources which Google has. Logical though, just keep it in mind, in case you forgot it. Mozilla’s power is truly the people. The devs and end-users. Only thing I suppose can help Fx right now, is the extremely good plan of impementations they got for Firefox. Already works perfectly. Fx 9 will be the King again almost in every aspect. I really praise Chrome’s existence. Made Mozilla to wake up and implement faster old ideas they had in their repository, eventually.

  3. Gueston says:

    bout time firefox! lets go!!

  4. DM says:

    Chrome beats firefox by a long way in one and a significant margin in two of those graphs

    while its good their improving end users wont see firefox 9 for many months and even then they still have a long way to go to catch up

    • Tiago Sá says:

      Firefox beats Chrome in sunspider.

      And Chrome beats Firefox in the benchmark called “Vet’s Vee Vich Vrowser Vis Vore Vimilar Vo Chrome” aka V8

  5. Guest says:

    Opera fanboys are so butthurt, it’s hilarious. Opera is uncompetitive and gets totally destroyed by Firefox in performance.

    http://www.conceivablytech.com/9229/products/a-year-in-browsers-dazzling-performance-gains/2

    Please note, 1600% slower than Firefox in performance. Typical Opera fanboys like Mr Jelly are clueless and ignorant. Firefox will only get faster and faster and there’s nothing Opera fanboys can do except keep using an irrelevant browser with 1% marketshare.

    • Rudi Visser says:

      Sorry, but are you retarded? That page you linked to shows Opera (in overall stats) as being 200% slower from the baseline than it is of Firefox 9.

      You took the results from ONE graph on HTML5 Hardware-Accellerated rendering rendering and determined that, based on this single graph, it’s 1600% slower? You do realise that there is no HW Acceleration in any of these builds which explains the “slowness”, as that is exactly what it’s testing. Cute, very cute ;-)

      PS. We never claim to have the fastest browser.
      PPS. We’re not “butthurt”.
      PPPS. I think most of the Opera users (or fanboys as you call them) are more mature than to care, really :-/
      PPPPS. You’re comparing a Firefox version 3 away from the current, to our “Next”.

      • Guest says:

        Opera fanboys are mad as usual. Enjoy your slow, outdated browser that is uncompetitive and is destroyed in performance.

      • apriorimeister says:

        Sure the 1600% is not fair, however it does show how slow Opera developers are at implementing something that all of the majors browsers have had for quite a while now. It also shows that chances are that Opera will likely never catch up with neither Firefox nor Chrome and possibly not even Safari after they implement WebKit2 at the rate that they are developing (I’m talking strictly in terms of performance here). Also, you can’t really blame the testers, because they were comparing the newest snapshots that were available to them. And if you think about it, Firefox 9 will probably come out at a somewhat similar time to Opera 12.

        • apriorimeister says:

          Sorry, it seems I was wrong about the Safari part as it seems that they have been using WebKit2 as of version 5.1.

        • Anonymous says:

          You are saying that Opera is slow at implementing everything everyone else is implementing, but that’s not very logical. That logic dictates that any given browser that hasn’t implemented everything everyone else has implemented is slow at implementing everything. Which means that all browsers are slow at implementing things because none of them implement everything all other browsers have implemented.

          • apriorimeister says:

            Hardware acceleration has been in development for a very very long time in Opera and the fact that they don’t have it implemented to this day is… well.. indicative of slow development – at least to me it is. Also in regards to your analysis, if the feature is objectively useful, well liked by the customers and has no real drawbacks then yes. Some of you may use the argument that HW Accel produces horrendous looking fonts and in turn has a real drawback, however that problem has mostly been fixed.

          • Anonymous says:

            Hardware acceleration is a great example.

            Opera’s rendering was so fast the other browsers had to use hardware acceleration to catch up. Because of Opera’s superior rendering performance hardware acceleration hasn’t really been necessary. Only now there are some better use cases for hardware acceleration, and lo and behold, Opera is on it.

            You are only confirming the point that browsers have different priorities. Opera’s priority has obviously not been to rush out hardware acceleration since the software rendering is already very fast.

            Not slow development, but rather a matter of prioritizing development resources.

          • apriorimeister says:

            “Opera’s rendering was so fast the other browsers had to use hardware acceleration to catch up.” – Sure it’s software rendering is fast, even slightly faster than that of Chrome. However rendering GPU accelerated layers using a GPU is orders of magnitude faster than of using a CPU. So frankly even a highly optimized software rendering engine is far from being a substitute for hardware acceleration – at least for me it isn’t. And many websites have and have had gpu accelerated layers for quite some time now.  I’ll repeat myself once again, they’ve been on it for a very very long time now and the results that we have seen from this: http://labs.opera.com/news/2011/02/28/ weren’t that amazing.

          • Anonymous says:

            You are missing the point. I said that Opera’s fast software rendering means that hardware acceleration is not and has not been critical. But it’s going to be important in the near future with more html5 games coming out.

            Your assumption is that hardware acceleration has been the #1 priority for Opera for all this time, and that is supposed to prove that they are “slow.” I pointed out that hardware acceleration clearly has not been their #1 priority so far because of the fast software rendering and lack of real content in need of hardware acceleration. Thus, disproving your claim.

            The existence of a proof of concept doesn’t mean that it’s their #1 most important thing to finish.

          • apriorimeister says:

            “.I said that Opera’s fast software rendering means that hardware acceleration is not and has not been critical” –  And I said that a fast software rendering engine is not a substitute for hardware acceleration, at least it is not for me, because I would much rather use that CPU processing power somewhere else. “Your assumption is that hardware acceleration has been the #1 priority for Opera for all this time” – Where and when did I say it was their number one priority? Exactly… I didn’t. They have a designated team working on it  and it has been on their agenda for a very long time now and from what we have seen is that the fruits or should I said fruit of their labor is frankly a bit rotten. I said that they are slow at developing and actually their desktop blog is proof of that, the number of bugs that they fix and the changes that they make is ridiculously small in comparison to chrome and firefox  so you really haven’t disproven my claim at all. Also, you don’t really have any proof what so ever that it has not been their 1# priority do you? You are just assuming and that is of course a very reasonable assumption, however it is just that – an assumption.

          • Anonymous says:

            I never said that software rendering is a substitute for hardware acceleration. I said that there hasn’t really been a need for hardware acceleration until now. That’s why it clearly hasn’t been a high priority. Web apps in need of hardware acceleration are few and far between right now.

            How do you know who’s working on what at Opera? You are just claiming this and that with no real arguments. I’ve been pointing out how HW acceleration hasn’t been necessary so far, so it’s logical that Opera hasn’t been putting a lot of resources into it yet.

            And basing your claims on heavily edited changelogs on a blog is even sillier. You do realize that not every single change is included, right? Just read the blog, and you’ll find that people report that bugs that are fixed aren’t even mentioned in the changelog for that build.

            Also, you are looking at the blog during vacation time in Norway. That’s even worse.

            Come on, logical arguments please. Not just FUD.

          • apriorimeister says:

            “I never said that software rendering is a substitute for hardware acceleration.” – Actually that’s exactly what you have been saying and you even said the same in your second and third sentences. Also, not just web apps can use accelerated compositing, I hope that you realize that. Yeah, I actually do, because they use job titles – http://twitter.com/#!/erikjmoller . “You do realize that not every single change is included, right? Just read the blog, and you’ll find that people report that bugs that are fixed aren’t even mentioned in the changelog for that build.” –  The only reason that you see that bugs which are not mentioned get fixed is because those bugs probably had a direct on indirect relation to the one that was fixed. If you have proof of the opposite – please by all means, post it here. Also, be my guest and look at the changelogs back when their were working fully staffed and see that my argument still stands on both feet. 

          • Anonymous says:

            No, that is not what I have been saying. What I’ve been saying is that there hasn’t been a real need for HW acceleration so far, but it will be more important in the future.

            Your argument fails because they have stated several times that the changelogs are edited, and they frequently post huge changelogs anyway.

          • apriorimeister says:

            “they have stated several times that the changelogs are edited” – Then please, be my guest and post a link to them saying so. The changelogs that they post are miniscule in comparison to the chrome and firefox ones.

          • apriorimeister says:

            “they have stated several times that the changelogs are edited” – Then please, be my guest and post a link to them saying so. The changelogs that they post are miniscule in comparison to the chrome and firefox ones.

          • Mikah says:

            IE9 has done a good job with Hardware acceleration but then its only developing for the one platform,  

            Heres a quote from the Opera Core Blog
            http://my.opera.com/core/blog/2011/02/28/webgl-and-hardware-acceleration-2

            “Our hardware acceleration is a bit different from what other browsers have implemented. Most of them do full hardware acceleration of all draw operations, but only on Windows Vista and Windows 7 – dropping to a more limited set of accelerated draw operations on other platforms. Our implementation will feature full acceleration on any OS with sufficient hardware support. This means we can also use fully hardware accelerated draw operations on Windows XP, Linux, Mac OS X and OpenGL ES 2 capable devices such as recent smart-phones and web-enabled TVs.

            OpenGL

            This build only has an OpenGL backend. That means your system must have an OpenGL 2.x compatible graphics card and related drivers for hardware acceleration and WebGL to work. In future builds we will also add a Direct3D backend, which will reduce the requirements on drivers and should work out of the box on most modern systems.”

            That was published in February 2011 theres a link where you can download Opera with the Open GL drivers & have a 1,000 fish zipping around at 60fps on the fishtank demo if your graphics card supports OpenGL.