Yes, though the loop unrolling is news to me. I'll have to keep that in mind next time I'm trying to squeeze some extra performance out of a loop. btw, found a static switch enhancement request here: https://issues.dlang.org/show_bug.cgi?id=6921