You are not logged in.

#1 2009-03-20 01:05:55

firecat53
Member
From: Lake Stevens, WA, USA
Registered: 2007-05-14
Posts: 1,542
Website

Firefox speed -- chromeexperiment.com

I was just messing around today and tried out some of the cool browser tricks at chromeexperiment.com on my Firefox (arch64 - firefox 3.0.7 and then firefox-pgo-beta 3.1b3).

I was pretty disturbed to discover that on my machine (hp dv6000 - amd turion dual core 1.6GHz), both firefox and firefox-beta performed worse (subjectively, but pretty dang obvious) than both firefox (3.1b3) and chrome did running in a Virtualbox XP on this same computer. I even tried disabling all the extensions, just to make sure. Both of the linux firefoxes exhibited a distinct pause in the 'browser-ball' experiment every second or so ... move-stop-move-stop. There was no smooth motion. The motion was a tiny bit slower in the virtualized browsers, but it was smooth. I had no other programs running at the time, and there was nothing obvious eating up CPU cycles.

Do I have something configured wrong?? Anyone else tried this yet? I'm not quite sure where to look for problems. Normally, video playback (flash, dvd,mpeg etc) is smooth, and I THOUGHT I had a pretty good overall browsing experience.

Thanks,

Scott

Edit: Well, it seems as if the culprit is my nvidia drivers. I just switched to the  sata_nv driver and, amazingly, it all worked great. I'm using the 177.82 drivers, because the 180 series cause frequent lockups (even after adding pci=nommconf to menu.lst per the nvnews forum).

Last edited by firecat53 (2009-03-20 02:11:00)

Offline

#2 2009-03-20 03:18:42

sand_man
Member
From: Australia
Registered: 2008-06-10
Posts: 2,164

Re: Firefox speed -- chromeexperiment.com

Well I'm glad you worked it all out but overall I'm not happy with the state of Firefox these days sad


neutral

Offline

#3 2009-03-20 04:22:09

firecat53
Member
From: Lake Stevens, WA, USA
Registered: 2007-05-14
Posts: 1,542
Website

Re: Firefox speed -- chromeexperiment.com

Well, I wouldn't say it's worked out .... I have to use the nvidia drivers because the sata_nv driver causes a hard lockup on resume from suspend to ram. Given that it's a laptop, that's a pretty critical factor for me.  I just wish I could sort out what's up with the nvidia drivers and those javascript graphics (well, and the lockups, too,with the newer drivers)....very weird! Firefox 3.1b3 is definitely faster than 3.0.7 and I never really liked Opera that much, but not intending to start another best browser discussion!! Just figure out whats up with my setup smile

Scott

Offline

#4 2009-03-20 04:36:39

skottish
Forum Fellow
From: Here
Registered: 2006-06-16
Posts: 7,942

Re: Firefox speed -- chromeexperiment.com

I tested the Browser Ball example on my system and while the ball had clean movement, the CPU usage was way up. With two child windows I was showing at times 60-70% CPU usage spread over two cores. Even when the ball stopped, the CPU usage stayed as high as 14% with no activity from me and nothing but FF open. This machine in an AMD 6400+ (3200MHz) processor with fast RAM, so I can imagine what would happen on your machine.

Offline

Board footer

Powered by FluxBB