Accession Number : ADA135794

Title :   The Extension and Application of Global Compaction Techniques for Horizontal Scientific Code.

Descriptive Note : Final 1 Sep 81-31 Aug 83,

Corporate Author : YALE UNIV NEW HAVEN CT DEPT OF COMPUTER SCIENCE

Personal Author(s) : Fisher,J A

PDF Url : ADA135794

Report Date : 08 Nov 1983

Pagination or Media Count : 8

Abstract : This report summarizes the work completed in applying trace scheduling to the Floating Point Systems AP164 array processor. This ongoing project focuses upon trace scheduling. The two main aspects of the proejct are the development of a trace scheduling compiler and to design a single instruction multiple data (SIMD) floating point machine that is optimized for trace scheduling. The aspect of this project which has been done under this ARO grant was the development of a trace scheduler for the AP164 attached processor. This architecture is the most common example of a SIMD machine for which the user writes code. The AP164 offers limited parallelism in that it has three alu's and a limited cross-bar.

Descriptors :   *Compilers, *Data processing, *Floating point operation, *Coding, Parallel processing, Global, Optimization, Scheduling, Computer architecture, Heuristic methods

Subject Categories : Numerical Mathematics
      Computer Programming and Software
      Computer Hardware

Distribution Statement : APPROVED FOR PUBLIC RELEASE