DMCA
A Reduction Semantics for Array Expressions: The PSI Compiler (1994)
| Citations: | 7 - 2 self |
BibTeX
@TECHREPORT{Mullin94areduction,
author = {Lenore Mullin and Lenore Mullin and Scott Thibault and Scott Thibault},
title = {A Reduction Semantics for Array Expressions: The PSI Compiler},
institution = {},
year = {1994}
}
Share
 |
 |
 |
 |
||
OpenURL
Abstract
High Performance Computing is not only concerned with the performance of a program on a very fast processor, but also high performance on many fast processors perhaps of different speeds and topologies. Scientific computation and modeling often require real time visualizations on high resolution graphics terminals. The data structure most widely used in scientific computation is the array. Arithmetics operations and permutations are often used where its associated algebra transcends many scientific disciplines. In order to achieve high performance computing, the interfaces with which people interact with computers must be rapid and accurate with the ability to move across many platforms and adapt quickly to change. If a programming language supports an algebra useful to scientists, the programming language, and subsequent compiler, must optimize the resources of a particular machine or machines. That is, the compiled program must allocate the least amount of memory and do the least amo...
Keyphrases
array expression psi compiler reduction semantics high performance computing scientific computation programming language data structure many fast processor many platform fast processor high resolution graphic terminal real time visualization compiled program high performance particular machine subsequent compiler arithmetic operation algebra transcends many scientific discipline different speed
