High performance of mathematical functions is essential to speed up sc
ientific calculations because they are very frequently used in scienti
fic computing. This paper presents performance of important Fortran in
trinsic functions on the fastest vector supercomputers. It is assumed
that a relationship between CPU-time and the number of function argume
nts given to calculate function values is linear, and speeds of a func
tion were measured using the parameters ($) over bar r(infinity), and
($) over bar n(1/2). The author also examines how the speed of the fun
ction varies with respect to the selection of arguments. The computers
tested in the present paper are Gray C9016E/162564, Fujitsu VP2600/10
, Hitachi S-3800/480 and NEC SX-3/14R.