Mobile Safari Benchmarks Show the Power of Nitro JavaScript Engine in iOS 4.3

With iOS 4.3, Apple improved Mobile Safari via the implementation of the Nitro JavaScript engine, which is supposed to more than double the performance of JavaScript. Just how much are these improvements compared to Mobile Safari on iOS 4.2.1?

German blog macnews.de posted results of tests using Geekbench 2, Guage Mathematical Tool, Google’s V8 benchmark and SunSpider. Are Apple’s claims of almost double the performance true?

Here were their test methods (translated):

Each test was performed ten times in succession, each represented the average values. This ensures that a single “outlier” does not distort the overall result. For all benchmarks were also the same conditions: The test device (an iPhone 4) was only the most necessary apps and a freshly installed iOS equipped, then it had to prove his skills.

Here are the results of each test (titles all translated via Google Translate):

General performance tests with Geekbench don’t show much of a difference…

..but Google’s V8 benchmark reveals a huge 2.47x performance increase:

SunSpider tests show Mobile Safari in iOS 4.3 loads more than 50% faster (time in milliseconds, smaller number is better):

Here are the SunSpider results in detail (smaller numbers are better), which illustrate the true performance boosts of the Nitro JavaScript engine:

So it looks like the promised performance of ‘more than double’ from Apple is true, according to these tests. For those on iOS 4.3, have you noticed a huge speed bump in Mobile Safari?

Thanks for the email Don!

[macnews.de]