You can see some benchmarks comparing them here:

http://arewefastyet.com/#machine=12&view=breakdown&suite=asm...

http://arewefastyet.com/#machine=12&view=breakdown&suite=asm...

FTL beats v8 on almost all of those.

All these benchmarks have "asmjs" in their names, which tells me they are not benchmarking idiomatic JavaScript, but JS generated by emscripten. If that's the case, they do not represent the performance of hand-written applications, which is the majority of all apps on the web.

It would be more interesting to see results for benchmarks such as DeltaBlue or Richards. These would correlate much better with gmail and hipmunk performance, as opposed to in-browser demos of Unreal Tournament.

I wish I knew if lower is better or higher is better on the first page (y axis is not described on the graphs).

Lower in the graphs is always better on awfy, which corresponds to lower numbers are better in all benchmarks, except for octane where higher numbers are better (the axis is reversed so still lower in the graph is better).

For me, chrome still beats this webkit and the latest firefox for deep learning mnist in the browser:

http://cs.stanford.edu/people/karpathy/convnetjs/demo/mnist....

I'm trying to figure out how to measure performance here. Both Chrome and Firefox seem to take about the same wall-clock time to go through 5000 examples. Should I be waiting longer? Measuring something else?