A quick and dirty JavaScript speed test of the new iPad 2 gives an early indication of how the tablet’s processor, operating system and HTML improvements pay off in terms of performance.
The speed test measured how fast the iPad 2’s Safari browser executed JavaScript. The test showed that iPad 2 runs it about four times faster than the original iPad. Part of that boost is due to the custom dual-core processor, Apple Inc.’s A5 chip. But a significant contribution is from software improvements Apple made in the iOS 4.3 firmware, which is new with the new iPad, and in the Safari browser.
The faster a browser can process JavaScript, the faster and smoother even complicated Web sites will appear to run on your tablet or smartphone. The test made use of the SunSpider JavaScript Benchmark, which is available online to anyone.
Apple unveiled the iPad 2 last week in San Francisco, with a companion even in London, U.K., where Jason Jenkins, who writes for CNET UK’s Crave gadget blog, had the brilliant idea of actually running a key benchmark test in the few minutes that he had the tablet in his hands.
[Jenkins’ original post at the CNET UK was picked up and referenced by a range of tech Web sites, such as here at The Unofficial Apple Webblog; but the original CNET UK post, at this writing, is no longer available. A cached version could still be found.]
“The good news for anyone thinking of buying the iPad 2 is that it did incredibly well in the test, outperforming all the rivals we have in the building to compare with it,” Jenkins reported.
The resulting number is in milliseconds, so the lower the number, the faster the JavaScript is being processed. The results are shown in the accompanying bar graph or in the graph still hosted at CNET UK.
The rivals included the Google Nexus S smartphone, the Samsung Galaxy Tab, both runningAndroid, and the Android browser. Nexus S with Android 2.3.2, took 6,128 milliseconds to run the benchmark; the Samsung device, 7,066; iPad 2 took 2,097 seconds.
The competitors also include the original iPad and an iPhone 4, both of which initially ran the current iOS 4.2 firmware. Jenkins later their firmware, to a beta version of iOS 4.3, and found the new software led to big performance gains for both devices in JavaScript processing. For example, the benchmark processing time on iPhone 4 dropped from 10,414 milliseconds to 4,151; on the original iPad, from 8,321 to 3,261.
Jenkins concluded that overall iPad 2 is about 1.5 times faster on the Web than the original tablet.
The 4.3 iOS version will be released for existing iPads, iPhone 4 and iPhone 3GS models on March 11. Users likely will see a noticeable improvement in Web performance, at least in someapplications.
Jenkins argues that this level of improved Web performance make buying the original iPad a very good deal, especially with the current discount now being offered by Apple. “Our numbers indicate that — providing you’re mainly going to use the iPad for Web browsing rather than gaming or videoconferencing — that’s a pretty nifty deal, as the performance difference between the two [iPads] is closer than Apple’s marketing might lead you to believe,” he writes.
Jenkins quick benchmark is not intended as a definitive analysis of how well iPad 2, and iOS 4.3 with the latest Safari actually handle Web content. That’s a complex field, depending on more than just the raw processing power of the computing hardware. Performance can be heavily influenced by the various components of inside the Web browser, by the Web server itself, and the way that browser and server interact.
With iOS 4.3, Apple has incorporated into Safari an updated Nitro (formerly Squirrelfish) JavaScript engine. The engine is the component that interprets and executes the actual JavaScript either in a browser or for a Web server). Apple has an aggressive program to implement HTML5 specifications in Safari, and refine and improve its Web performance.
