IE9 vs Chrome 10 vs Firefox 4 vs Opera 11.01 vs Safari 5
Here are the browsers that will be held:
- Internet Explorer 9 (9.0.8112.16421) 32-bit
- Internet Explorer 9 (9.0.8112.16421) 64-bit Firefox 4 BC
- Chrome 10.0.648.133
- Safari 5.0.4
- Opera 11.01
- SunSpider JavaScript 0.9.1 – A benchmark developed by Mozilla JavaScript with a focus on the real world problem.
- V8 Benchmark Suite – A pure JavaScript benchmarks by Google in the height of the V8 JavaScript engine used.
- Peacekeeper – Futuremark JavaScript exam stress tests have features such as animation, navigation, forms and other objects commonly used.
- Kraken 1.0 – Another JavaScript benchmark developed by Mozilla. This is based on the SunSpider, but has a number of improvements.
All tests on the machine Windows 7 64-bit running Q9300 quad-core 2.5GHz processor with 4GB RAM and NVIDIA GTX 260 graphics card done.
Ok, with the loser out of the way, what the profit of the AOS-browser? Well, obtained for all four benchmarks, Chrome 10, two 32-bit IE9 and won one for Firefox 4 BC, in the strict sense, chromium 10, the winner. But SunSpider tests is so little separation of the four fastest browser (a shade over 40 milliseconds) that can almost say that it is authentic. Then, AOS V8 test, Chrome 10th Aces, but according to the AOS given that Google is expected by his own admission, to vote in the benchmark is that the JavaScript engine V8.
So what, AOS completed? Is simple, 64-bit IE9 shockingly bad, and the other browsers, on the whole fairly evenly matched. I am pleasantly surprised that aces Aom IE9 actual 32-bit SunSpider test, but I hope that the coming weeks, Google will succeed in coming forward again (Microsoft, RBA painfully slow cycle update for IE (especially Google, AOS aggressive update cycle comparison) works against the long term. Microsoft has hard work on IE, the slowest in the pack with one of the fastest. That, however, it is worth recognizing.
Bottom line, I don really, I think, AOT test JavaScript performance problem, and certainly also in real-life, AOS is difficult to differentiate between the browsers (5 despite some HTML pages, see there is quite a specific Browser optimized). In fact, unless management of the parties is to support JavaScript performance by an order of magnitude, shaving a few milliseconds here and there hardly matter anymore.
