Double precision exponentiation is performed in the same manner as single-precision exponentiation by means of two routines: RTOD (real-to-double) and DTOD (double-to-double). In either case, the result is double-precision.