Javascript Performance Thread

Discussion about official Mozilla Firefox builds
Post Reply
User avatar
bogas04
Posts: 977
Joined: May 18th, 2010, 1:14 am

Javascript Performance Thread

Post by bogas04 »

I think the time has come when this is to be discussed thoroughly
Last edited by bogas04 on December 27th, 2011, 12:23 am, edited 2 times in total.
bogas04.github.io
MacBook Air Mid 2013 |@bogas04
Harsh86
Posts: 94
Joined: August 23rd, 2004, 4:34 am

Re: Firefox 4.0 TM and JM Performance Thread

Post by Harsh86 »

Brendan Eich mentions in this bug https://bugzilla.mozilla.org/show_bug.cgi?id=578517 that Andreas Gal has a pretty big TM win brewing. Does anyone know which bug he's talking about?
Solidstate89
Posts: 42
Joined: July 4th, 2010, 12:02 pm

Re: Firefox 4.0 TM and JM Performance Thread

Post by Solidstate89 »

I'm getting fairly faster numbers in Sunspider since the latest Nightly. On my laptop I usually got around ~700ms in the test, and now I get around ~600. One of the tests I ran gave me a 580.
Josa
Posts: 7360
Joined: July 28th, 2009, 4:52 pm

Re: Firefox 4.0 TM and JM Performance Thread

Post by Josa »

KriTenKs wrote:Brendan Eich mentions in this bug https://bugzilla.mozilla.org/show_bug.cgi?id=578517 that Andreas Gal has a pretty big TM win brewing. Does anyone know which bug he's talking about?

Maybe
https://bugzilla.mozilla.org/show_bug.cgi?id=577882 and https://bugzilla.mozilla.org/show_bug.cgi?id=577883?
7% is a big win for TM these days.
iwod
Posts: 1033
Joined: July 18th, 2003, 10:09 pm

Re: Firefox 4.0 TM and JM Performance Thread

Post by iwod »

Josa wrote:
KriTenKs wrote:Brendan Eich mentions in this bug https://bugzilla.mozilla.org/show_bug.cgi?id=578517 that Andreas Gal has a pretty big TM win brewing. Does anyone know which bug he's talking about?

Maybe
https://bugzilla.mozilla.org/show_bug.cgi?id=577882 and https://bugzilla.mozilla.org/show_bug.cgi?id=577883?
7% is a big win for TM these days.


Arh,. I hope not when chrome / Opera are 100% faster then us.
User avatar
sciguyryan
Folder@Home
Posts: 2181
Joined: November 10th, 2004, 1:33 pm
Location: Wales

Re: Firefox 4.0 TM and JM Performance Thread

Post by sciguyryan »

iwod wrote:Arh,. I hope not when chrome / Opera are 100% faster then us.


Not really.

By the way. It should be noted on this thread that Are We Fast Yet? is the site that should be checked for the most up-to-date statistics on the WIP JS engines :)
Cheers!

Ryan Jones
User avatar
bogas04
Posts: 977
Joined: May 18th, 2010, 1:14 am

Re: Firefox 4.0 TM and JM Performance Thread

Post by bogas04 »

He is right to some extent , they really are twice as fast as us , but then we don't have a good benchmark platform , and they can also cheat by favouring a benchmark etc
bogas04.github.io
MacBook Air Mid 2013 |@bogas04
User avatar
sciguyryan
Folder@Home
Posts: 2181
Joined: November 10th, 2004, 1:33 pm
Location: Wales

Re: Firefox 4.0 TM and JM Performance Thread

Post by sciguyryan »

bogas04 wrote:He is right to some extent , they really are twice as fast as us , but then we don't have a good benchmark platform , and they can also cheat by favouring a benchmark etc


You cannot only work from benchmarks. Benchmarks do not represent real world cases in most cases, actually they are usually tailored to meet the specific performances of the browser vendor that built them. Mozilla and Opera are two of the honest ones who did not go off and build their own, even though I am fairly sure Mozilla could easily write one on which their tracer bested everything easily.

Benchmarks are a really, really bad way of representing speed but those are all we have unfortunately :(
Cheers!

Ryan Jones
mat--
Posts: 669
Joined: May 20th, 2010, 8:12 am

Re: Firefox 4.0 TM and JM Performance Thread

Post by mat-- »

Also, fatvals is supposed to improve the way the engines talk to each other, not the engines itself, so it might not show improvements on benchmarks but it will feel much better on real world scenarios.
mat--
Posts: 669
Joined: May 20th, 2010, 8:12 am

Re: Firefox 4.0 TM and JM Performance Thread

Post by mat-- »

Josa wrote:
KriTenKs wrote:Brendan Eich mentions in this bug https://bugzilla.mozilla.org/show_bug.cgi?id=578517 that Andreas Gal has a pretty big TM win brewing. Does anyone know which bug he's talking about?

Maybe
https://bugzilla.mozilla.org/show_bug.cgi?id=577882 and https://bugzilla.mozilla.org/show_bug.cgi?id=577883?
7% is a big win for TM these days.

There is also https://bugzilla.mozilla.org/show_bug.cgi?id=564953 which gives a 12% speedup on 32bit and 25% on 64bit builds.
User avatar
_Alexander
Posts: 1197
Joined: April 1st, 2010, 2:24 pm
Location: Your augmented reality

Re: Firefox 4.0 TM and JM Performance Thread

Post by _Alexander »

RyanJ wrote:
bogas04 wrote:He is right to some extent , they really are twice as fast as us , but then we don't have a good benchmark platform , and they can also cheat by favouring a benchmark etc


You cannot only work from benchmarks. Benchmarks do not represent real world cases in most cases, actually they are usually tailored to meet the specific performances of the browser vendor that built them. Mozilla and Opera are two of the honest ones who did not go off and build their own, even though I am fairly sure Mozilla could easily write one on which their tracer bested everything easily.

Benchmarks are a really, really bad way of representing speed but those are all we have unfortunately :(

They are really really good way to win over users from using other browsers though.
http://magneticpudding.com/ <- My Blog
i5 3570k @ 4.5 Ghz / NV 660 / 32GB DDR3 / 1080p LCD / SSD (120 + 180) / W8 ||| Atom N270 / NV ION / 3GB DDR3 / SSD / 1366x768 / W8
User avatar
Tiago Sá
Posts: 110
Joined: June 14th, 2010, 6:16 am

Re: Firefox 4.0 TM and JM Performance Thread

Post by Tiago Sá »

Do we really care about that? It's not like they're leaving for Internet Explorer, so that's good. And it's not like Mozilla Corporation is in financial troubles...
Isn't it annoying when someone uses an avatar that says absolutely nothing about themselves and is there just to look cool?
Isn't it annoying when it
doesn't look cool?
User avatar
patrickjdempsey
Posts: 23686
Joined: October 23rd, 2008, 11:43 am
Location: Asheville NC
Contact:

Re: Firefox 4.0 TM and JM Performance Thread

Post by patrickjdempsey »

Hera wrote:They are really really good way to win over users from using other browsers though.


Which is exactly why Microsoft and Google have their own and also make reports all of the time full of spurious claims about speed and usage statistics. The whole mess reminds me of claims you see in car commercials. Everyone builds the fastest, safest, roomiest, most efficient cars and win the most awards in their own tests and commercials.
Tip of the day: If it has "toolbar" in the name, it's crap.
What my avatar is about: https://addons.mozilla.org/en-US/seamonkey/addon/sea-fox/
mat--
Posts: 669
Joined: May 20th, 2010, 8:12 am

Re: Firefox 4.0 TM and JM Performance Thread

Post by mat-- »

https://bugzilla.mozilla.org/show_bug.cgi?id=579140 is interesting, not sure if they will merge TM to m-c when this is reviewed.
trieste
Posts: 618
Joined: September 4th, 2003, 4:15 am
Location: Singapore

Re: Firefox 4.0 TM and JM Performance Thread

Post by trieste »

Hera wrote:They are really really good way to win over users from using other browsers though.

Does Mozilla really need, or want, these users? I'm not making an elitist reference, but wondering about the users who are easily swayed by marketing claims. These are the ones who ought to stay on Chrome simply because Chrome is not customizable (no about:config to mess up).
Post Reply