How to improve the performance of programs calling mathematical functions

Taking advantage of IBM XL C/C++ or XL Fortran compiler auto-vectorization

From the developerWorks archives

Robert Enenkel and Daniel Zabawa

Date archived: December 20, 2016 | First published: April 13, 2010

This article introduces the IBM MASS high-performance mathematical libraries, and demonstrates how to benefit from them — without the need for source program changes — by using the auto-vectorization capability of the IBM® XL C/C++ and XL Fortran compilers. After introducing the concept of auto-vectorization and the associated compiler options, a case study of a discrete Fourier transform program is offered as a real life example of auto-vectorization. Timing results demonstrate that speedups of up to 8.94 times are obtained by the compilers on the example program, via the automatic invocation of MASS by auto-vectorization.

This content is no longer being updated or maintained. The full article is provided "as is" in a PDF file. Given the rapid evolution of technology, some steps and illustrations may have changed.



static.content.url=http://www.ibm.com/developerworks/js/artrating/
SITE_ID=1
Zone=Rational, Multicore acceleration
ArticleID=481621
ArticleTitle=How to improve the performance of programs calling mathematical functions
publish-date=04132010