Mozilla Asks, 'Are We Fast Yet?'

The above charts show the performance of JavaScript engines across different architectures. The tests shown are the common SunSpider and V8 JavaScript benchmarks, with output measured in milliseconds. The tests are run once a day, and the graphs show the last five weeks or so of results. Go to the real site and click on […]

The above charts show the performance of JavaScript engines across different architectures. The tests shown are the common SunSpider and V8 JavaScript benchmarks, with output measured in milliseconds. The tests are run once a day, and the graphs show the last five weeks or so of results.

Go to the real site and click on all the clicky bits.

The green line is Google V8, the red line is Apple Nitro, and the orange and black lines are Mozilla's two engines, JaegerMonkey and TraceMonkey, respectively. The purple lines reflect Mozilla's new approach of running the engines concurrently. As you can see, it speeds things up.

But the answer to the question being asked by the URL is "No" – Google is currently either on par with Apple Safari or slightly better, depending on the test and the architecture. Mozilla is improving, but still has a lot of catching up to do.

This testing tool is maintained by Mozilla's JavaScript team. I found out about it earlier today when John Resig, the guy behind jQuery and a Mozilla employee, tweeted the link. It's an effective motivational tool, especially since it shows how slow Mozilla's engines were only a month ago, and how quickly the team is gaining on the leaders.

A couple of caveats: The tests aren't run in the browser, they are run from the command line. Also, a Mac Mini in doing the testing, so Internet Explorer isn't represented. From what we've seen of IE9's pre-release code, the browser is incredibly fast. We're curious to see how its JavaScript engine stacks up.

Also, no Opera. Opera's Carakan engine is also blazing fast, but it's not represented here.

Check out the page's FAQ for more details. Also, the code for the test is open source, so if you have philosophical issues with these methods, build your own testing environment.

Update: Here's a much more detailed post about Mozilla's performance on JavaScript benchmarks by Rob Sayre.

See Also: