rusha

It's definitely not "just a compiler target," not more than the assembly language in which people still write some code by hand when there's need for that. The real world example: https://github.com/srijs/rusha (see the benchmarks -- Firefox 21 uses asm.js) and the source https://github.com/srijs/rusha/blob/master/rusha.js I don't see that the code got ugly. The calculations needed are the same, the most obvious differences from the "non-asm.js" code is the "asm.js" and a few |0 marks. And some of such |0 constructs can improve speed even on the browsers which don't do full "asm.js". So the engines will get smarter by recognizing such constructs even without the full support of "asm.js" segments. It's a win-win already.

What I meant to say was not to use asm.js as target for compilation, but rather implementing performance-critical module in asm.js subset and get the extra performance out of V8. Sort of like how Rusha[1] did it. [1]: https://github.com/srijs/rusha/commit/e435ccdd709b6a3374879c...

Cool. I recently wrote a low-level sha1 implementation in pure javascript, trying to leverage asm.js (http://github.com/srijs/rusha). Can you tell me, how did you confirm you wrote valid asm.js and there was no fallback-mechanism kicking in? I benchmarked my implementation, of course, and noticed a 2x speedup. However, this seems quite less for the this type of low-level algorithm...