100,000k with the most efficient methods I could come up with (also added Fortran, which is tricky because the standard library only supports 33-digit precision) :