Implementing pipelined computation and communication in an HPF compiler
Many scientific applications can benefit from pipelining computation and communication. Our aim is to provide compiler and runtime support for High Performance Fortran applications that could benefit from these techniques. This paper describes the integration of a library for pipelined computations in the runtime system. Results on some application kernels are given.
Unable to display preview. Download preview PDF.
- [BD96]T. Brandes and F. Desprez. Implementing Pipelined Computation and Communication in an HPF Compiler. Technical report, LIP-ENS Lyon, 1996.Google Scholar
- [BZ94]Th. Brandes and F. Zimmermann. ADAPTOR — A Transformation Tool for HPF Programs. In K.M. Decker and R.M. Rehmann, editors, Programming Environments for Massively Parallel Distributed Systems, pages 91–96. Birkhäuser, April 1994.Google Scholar
- [DT92]F. Desprez and B. Tourancheau. LOCCS: Low Overhead Communication and Computation Subroutines. Technical Report 92-44, LIP-ENS Lyon, December 1992.Google Scholar
- [Kin88]C. King. Pipelined data parallel algorithms: Concept, design and modeling. PhD thesis, Michigan State University, Department of Computer Science, 1988.Google Scholar
- [KLS+94]C. Koelbel, D. Loveman, R. Schreiber, G. Steele, and M. Zosel. The High Performance Fortran Handbook. The MIT Press, Cambridge, MA, 1994.Google Scholar
- [Tse93]C. Tseng. An optimizing Fortran D compiler for MIMD distributed-memory machines. PhD thesis, Rice University, Houston, Texas, 1993.Google Scholar