让人很无语的WorkPress，用表格把这个页面都撑开了。用截图吧。。。

All tests were on the same system (2Ghz AMD Athlon XP2400+ Windows XP 1.5Gb RAM)
All times are in milliseconds
All tests were repeated multiple times on different days (after reboots) and on different systems to ensure that the results were consistent, but variation is inevitable. Try the tests on your own system.. and please report problems..
Flash tests (FP7,8,and 8.5) were loaded in IE6
Flash Players used: 7.0.19.0,  8.0.5.0,  8.5.0.133,  9.0.18.60
* indicates first run / second run of the same function
Run the Java version (prepared by Bill Riemers) of the test here (note.. it takes roughly 70 seconds for the applet to load the first time on my system. Results shown above are from reloading the page) The detailed result from running his benchmark on my system is here (yep.. I rounded some numbers.. we’re comparing 4649ms vs 6ms here)
Tests run on 11/24/06 used updated Flash SWF from Ted Patrick and used the released Flash Player 9. Some of the other tests (JS and FP7) were run again just to check that the results were still in line since they were run in the Fall of 2005. I didn’t rerun all of the JavaScript tests for IE6, and didn’t change the results because they were only off by a few miliseconds on most of them. I added the results for running JS in Opera.. and reran the Firefox tests since a new version was released.
Note that I did not run these test in the “stand alone” Flash Player. They’re run in the browser because that’s where almost all of my Flash code is executed. However, I’ve found in the past that code generally executes faster in the stand alone player.
1. JavaScript Speed test #3
2. JavaScript MD5 implementation
2. Flash implementation other JS tests were based (sometimes loosely) on some of these
Download sources:
HTML,
FLAs/SWFs (ActionScript 2)
MXML (original) * note that my MXML is quite crappy because I hadn’t even started learning ActionScript3 when I put this together..
MXML by Brian Deitte (strict typing made huge improvements over my original code)
MXML by Ted Patrick (note that all tests are run sequentially without unloading the SWF.. results vary greatly when I test multiple times)
Java by Bill Riemers

Copy From :http://oddhammer.com/actionscriptperformance/set4/

虽然最近已经没搞Java了，但是绝对不会忘记他，即使那本厚厚的《Thinking in Java 4nd》离读完还遥遥无期~