Which Web Browser is King?

It’s a highly debatable topic, one that tends to be subjective and fraught with inconsistencies. For example, latency on the Internet can dramatically affect browser speed. One day, Google Chrome can load IGN.com faster than butter on a banana, the next day (say, when a new Gears of War 2 review posts), latency can slow the site to a crawl, and Chrome seems like a dud. That’s why, when you see speed tests for browser that claim “Chrome loads faster” it’s important to ask a few questions: loaded when, over what broadband speed, with what other apps running, on what machine?

click on image for full view

Speed, of course, isn’t everything. We ran into an interesting compatibility glitch or two completing some of our tests. If a browser can’t finish a particular benchmark, it doesn’t get a score. 

What does it mean, ultimately? For the average user, it may not mean much. But for those of us who spend vast amounts of time on the Internet, for work and play, browser performance is a big issue. Waiting for sites to load is no fun when you’re trying to get work done.

 Testing Methodology

In this comprehensive browser speed test, I pulled out all of the stops. First, I found just about every speed test site around. Google offers a V8 test that tests JavaScript speed for Web apps. I also found test sites for SVG (scalable vector formats—an XML specification), Acid3 (a Web standards test engine), Canvas (an HTML extension), and DOM (document object model—an XML object). And, I found a great Flash test site that loads crabs into the browser one by one until the frame rate drops below 25FPS—it was originally designed for the PlayStation 3 but it works great as a generic speed tester.

Of course, I also tested ten content-rich sites for pure load time. Again, latency plays a major factor here, so I also tested all the sites several times through the day and used the average time.

I didn’t stop there. I repeated all of these tests on three different computers. I performed the tests on a Dell XPS M1530 laptop with 4GB of RAM, a 2.50GHz processor (the Intel T9300), and Windows Vista Home Premium 32-bit as the base machine—a typical laptop. (Those are the results you see below.)

To verify the benchmark sites, I repeated the tests on a high-end workstation: an HP xw6400 with 16GB of RAM, a 3.0GHz processor, and Windows Vista Ultimate 64-bit. Finally, I used a slower-than-dirt Mac Mini Intel Single Core 1.5GHz Mac with 2GB of RAM as a third sanity test. Since latency is an obvious factor, I tested over the same 2Mbps broadband feed in my home.

It’s interesting to note that the high-end PC and the low-end Mac Mini did not produce different results in any of these tests, with only slightly different rendering results (a second or two) for site loading and JavaScript. That means, the important issue is broadband speed and the browser itself and how it handles the code.

Now, a note about which browsers I used. For all tests, I used the version that is widely available to most users, not an obscure point version beta. (Google Chrome, which is only available as a beta, is the exception to this rule—I used version 0.3.154.9.) If you run all of these tests yourself on the Firefox 3.1 beta, post your results in the discussions—we’d like to see them. However, when we contacted Mozilla, they advised us to use the release version which is the most compatible and bug-free. (The latest 3.1 beta does not include the code that optimizes JavaScript apps, however.)

Otherwise, I used Firefox 3.04, the Chrome beta, the release version of Internet Explorer 7 (7.0.6001.18000), Opera 9.6, and Safari 3.2 (526.26.13) for each round of testing. If you have any concerns or comments about the results, post in the discussions—however, suffice it to say, at the time of this writing, under the same circumstances, I tested each browser on the same day to make sure they were all measured under the same conditions without any biases.

After each test run, I also assigned an arbitrary score, based on a ten point scale, with the winner getting ten points. At the end, all the scores are rolled up into a single score.

In the end, if you’re like me, you want to know the bottom line first. Sure, the details might be interesting, but you just want to know—at this point in time, which browser runs the fastest? Which one will help you get your work done faster than the others? And which one is the best for complex Web applications like Zoho Writer or Tokbox.com? That prize goes to Google Chrome, a browser that excelled at just about every benchmark, especially in the ten Web site speed tests.