As for general use of kotlin in high performance c...
# mathematics
a
As for general use of kotlin in high performance computations, there is indeed a boxing problem, which could not be currently solved without ugly solutions, but it won't stay this way. On one hand, JVM moves in direction of better escape analysis. I already told about Graal results. Also we will have Valhalla sooner or later. On second hand, it is not very complicated to add primitive specialization or some kind of intrinsic to avoid boxing. Kotlin team is willing to help, but first we need to understand what are the most critical moments.