Volume 7, Issue 1 -
Spring/Summer 1999

Volume 6, Issue 3
Fall 1998

Volume 6, Issue 2
Spring/Summer 1998

Volume 6, Issue 1
Winter 1998

Volume 5, Issue 4
Fall 1997

Volume 5, Issue 3
Summer 1997

Volume 5, Issue 2
Spring 1997

Volume 5, Issue 1
Winter 1997

Volume 4, Issue 4
Fall 1996

Volume 4, Issue 3
Summer 1996

Volume 4, Issue 2
Spring 1996

Volume 4, Issue 1
Winter 1996

Volume 3, Issue 4
Fall 1995

Volume 3, Issue 3
Summer 1995

Volume 3, Issue 2
Spring 1995

Volume 3, Issue 1
January 1995

Volume 2, Issue 4
October 1994

Volume 2, Issue 3
July 1994

Volume 2, Issue 2
April 1994

Volume 2, Issue 1
January 1994

Volume 1, Issue 4
October 1993

Volume 1, Issue 3
July 1993

Volume 1, Issue 2
April 1993

Volume 1, Issue 1
January 1993

CRPC Researchers Use UTCOMP Reservoir Simulation Code to Test Fortran D

A collaborative project between the Parallel Compilers group of the CRPC, directed by Ken Kennedy, and the Flow in Porous Media group, directed by Mary Wheeler, is investigating whether a data-distribution language (such as Fortran D or HPF) can express the parallelism in applications well enough to enable compilation systems to generate efficient parallel code. Uli Kremer and Marcelo Rame are leading this collaboration. Fortran D was tested on a small section of a widely used reservoir simulation code (UTCOMP, produced by the Petroleum Engineering Department at the University of Texas, Austin). The original code had been optimized for vector supercomputers by having most of the arrays "linearized" (declared as long vectors). Unfortunately, this machine-dependent programming style inhibits many compiler optimizations for parallel machines. The results provided insight into the effectiveness of Fortran D and its current prototype compiler. In addition, the study shed light on the programming style that should be used to make programs portable across a variety of parallel machines.

A number of typical procedures from UTCOMP were rewritten in Fortran D and then translated by the Fortran D compiler. The generated program was run on an Intel iPSC/860 distributed-memory multiprocessor. The results indicate that the Fortran D language and compilation system are well suited to express the parallelism in reservoir modeling computations, provided that the code is written in a data-parallel style. Speedups for the selected parts of UTCOMP were highly encouraging. These results will also be used to help design enhancements to Fortran D and its compilation technology.


Table of Contents