Damn, 50x better =O

Acutally it can be a lot more if you are talking about pause latency, I've seen >200x improvements on max pause time on some syntethic benchmarks

Really exciting stuff, definitely gonna be running some benchmarks on this this week!

Definetly interested to hear your results. The synthetics suggest that it is just winning in everything except raw allocation performance, but unless you're allocating millions of objects per second then that shouldn't matter.