Performance issues of scientific programming in Ada 95

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

1 Scopus citations

Abstract

Motivate by the increasing size and complexity of scientific software, we consider the benefits of Ada 95 versus Fortran and C++. Using a sample problem derived from a recent evaluation of C++ [4], we test the performance of a series of Ada 95 implementations designed to supply an elegant interface for scientific programming. We test two compilers, GNAT and PowerAda, on a single processor of the IBM SP2 at the Ohio Supercomputer Center, and we find that the Ada 95 implementations take between 1.5 and 4 times as long as hand- coded Fortran and C implementations. We analyze how all the implementations vary in performance, usability, and general applicability.

Original languageEnglish
Title of host publicationProceedings of the Conference on TRI-Ada 1997, TRI-Ada 1997
EditorsSusan Carlson
PublisherAssociation for Computing Machinery, Inc
Pages279-293
Number of pages15
ISBN (Electronic)0897919815, 9780897919814
DOIs
StatePublished - Nov 1 1997
Externally publishedYes
Event1997 Conference on TRI-Ada, TRI-Ada 1997 - St. Louis, United States
Duration: Nov 9 1997Nov 13 1997

Publication series

NameProceedings of the Conference on TRI-Ada 1997, TRI-Ada 1997

Conference

Conference1997 Conference on TRI-Ada, TRI-Ada 1997
Country/TerritoryUnited States
CitySt. Louis
Period11/9/9711/13/97

Fingerprint

Dive into the research topics of 'Performance issues of scientific programming in Ada 95'. Together they form a unique fingerprint.

Cite this