Abstract: The vectorization of loops invoking math function is an important optimization that is available in most commercial compilers. This paper describes a new command line option, -fsimdmath, ...