As I tell everyone when it comes to browsers: "It is the dealers choice"
I prefer Firefox and Internet Explorer. I use Firefox for everything, and if I encounter a site that is just not working (properly) on Firefox, I will launch IE. Otherwise, IE is never really used.
I've been playing with a few test systems for the last week or so. 5 in total, all Windows 7 Home Premium with a Core 2 Duo (2.4GHz) and 8GB of RAM. 2 of them have Chrome, 2 of them have Firefox, and 1 is my IE control system. They are installed with the same exact software (Java, Adobe, Flash, etc.) for the browsers. What I do with one, I do with the others such as same web pages for all. Needless to say, my wife is getting tired of some of my at-home projects, but not this one, she finds it fun and has helped me out a bit. All browsers being the latest release.
We use between 3 and 8 tabs, depending on what I am trying to achieve.
Baseline (3 Tabs) - Pandora (Not Playing), Facebook, Google (no search yet).
- Firefox averages 325MB of RAM consumption (1 process)
- Chrome averages 350MB of RAM consumption (8 processes)
- IE averages 330MB of RAM consumption (1 process)
Light Stress Test (5 Tabs) - Pandora (Not Playing), YouTube (1 Video), Facebook, Google (search), Wikipedia
- Firefox averages 440MB of RAM consumption (1 process)
- Chrome averages 728 of RAM consumption (12 processes)
- IE averages 500MB of RAM consumption (1 process)
Heavy Stress Test (12 tabs) - Pandora (Not Playing), YouTube (1 Video), Facebook, Google (3 search tabs), Wikipedia (2 tabs), 2xPDF's or downloads varying sizes, Gmail (1 tab, viewing email), random web page
- Firefox averages 700MB of RAM consumption (1 process)
- Chrome averages 900MB of RAM consumption (18 processes)
- IE averages 750MB of RAM consumption (1 process)
What I learned from this experiment:
- Under the baseline, things were still smooth across all 5 systems. The chrome systems however had a higher CPU utilization throughout, mainly because the CPU had more processes to process.
- Under the light stress test, Firefox ran like a champ and performed amazingly well with switch tabs, changing pages, etc. Chrome systems ate up the CPU like it was nothing, and provided in many cases a slower response compared to Firefox (visual). IE, while it performed well for the baseline, it started getting clunky and slow to respond and load pages, worse than Chrome was. When we had the browsers launch the above pages as the default homepages (all 5 tabs at once), Firefox performed the best, Chrome was moderate though some tabs crashed stalling the whole system, and IE was slow assuming it didn't crash.
- Under the Heavy stress test, the CPU for all 5 systems was strained at times. Firefox performed very well, though occasionally Flash would crash, but a reload of the web page fixed that right up. When flash would crash on Firefox, the whole browser became unresponsive from 30 seconds to 5 minutes but it never fully crashed, it just reported that flash did. Chrome was way to heavy on the systems, keeping the CPU utilization consistently above 60% (Firefox would stabilize at 20% after loading, IE around 40% consistently). Chrome was heavy, slow, and the random crashes of random tabs got annoying very quick. Chrome was my least favorite in this test. IE was slow as well, and though it crashed often, or would freeze for minutes on end, recovery was a breeze compared to Chrome as the CPU wasn't straining as bad to open IE compared to Chrome.
My least favorite of the tests overall had to be IE. Slow opening, slow loading, slow, slow, slow. Might be great for some things, but not when you have multiple tabs open.
Chrome was horrid when it came to the CPU. All the processes fighting over the CPU, making sure it stayed super busy. While it was nice that when a tab crashed, it was only that tab, it could take minutes before the notification appeared for that tab because the CPU was just under so much stress. Would I switch to chrome? NOPE! What started out as a test of RAM consumption quickly turned in to a battle of which browser is nicer to the CPU.
Firefox had its drawbacks, such as when a tab crashed, the whole thing crashed. However, it was quicker to recover, CPU utilization not as bad, and consistently used the least amount of RAM out of all of them. Compared to Chrome, Firefox was smoother, faster, and responded a lot better than Chrome and IE.
When doing downloads, consistently Chrome just downloaded without asking if I wanted to open or save. Very annoying and worrying. I treated this test like a casual user, so I didn't modify any settings in the browser. Chrome and IE built a massive temp storage on the system over this time, over 4.5GB of Temp files. The Firefox systems were both sitting around 3.8GB of Temp Files. Remember, all 5 went to the same websites, all 5 downloaded the same things. IE didn't surprise me, but Chrome was worrying with how much it saves and how it was handling downloads.
