Optimizing performance on parallel hardware is an iterative process. The flow chart shows the various steps involved.
This article describes how to use Intel® Advisor and VTune™ Amplifier in a Linux* distributed environment. While specifically designed to collect performance data at the node and core level, both tools can be used with MPI. The document covers basic utilization with an MPI launcher from the command line, as well as more advanced customizations.
Subroutine dummy arguments can be pointers or assumed shape arrays, e.g.:
SUBROUTINE SUB(X, Y) REAL, DIMENSION(:) :: X ! assumed shape array REAL, DIMENSION(:), POINTER :: Y ! pointer
This avoids the need to pass parameters such as array bounds explicitly. The Fortran standard allows the actual arguments to be non-contiguous array sections or pointers, e.g.:
This page provides the current Release Notes for Intel® Parallel Studio XE. The notes are categorized by year, from newest to oldest, with individual releases listed within each year.