A comparison of PETSc library and HPF implementations of an archetypal PDS computation 1 This work w

Two paradigms for distributed-memory parallel computation that free the application programmer from the details of message passing are compared for an archetypal structured scientific computation --- a nonlinear, structured-grid partial differential equation boundary value problem --- using the same algorithm on the same hardware. Both paradigms, parallel libraries represented by Argonne''s PETSc, and parallel languages represented by the Portland Group''s HPF, are found to be easy to use for this problem class, and both are reasonably effective in exploiting concurrency after a short learning curve. The level of involvement required by the application programmer under either paradigm includes specification of the data partitioning (corresponding to a geometrically simple decomposition of the domain of the PDE). Programming in SPMD style for the PETSc library requires writing the routines that discretize the PDE and its Jacobian, managing subdomain-to-processor mappings (affine global-to-local index mappings), and interfacing to library solver routines. Programming for HPF requires a complete sequential implementation of the same algorithm, introducing concurrency through subdomain blocking (an effort similar to the index mapping), and modest experimentation with rewriting loops to elucidate to the compiler the latent concurrency. Correctness and scalability are cross-validated on up to 32 nodes of an IBM SP2.

[1]  Guoliang Xue,et al.  The MINPACK-2 test problem collection , 1992 .

[2]  David E. Keyes,et al.  Domain-Based Parallelism and Problem Decomposition Methods in Computational Science and Engineering , 1995 .

[3]  David E. Keyes,et al.  On the Interaction of Architecture and Algorithm in the Domain-based Parallelization of an Unstructu , 1997 .

[4]  William Gropp,et al.  Parallel Implicit PDE Computations , 1997, Parallel CFD.

[5]  Stamatis Vassiliadis,et al.  Parallel Computer Architecture , 2000, Euro-Par.

[6]  John E. Dennis,et al.  Numerical methods for unconstrained optimization and nonlinear equations , 1983, Prentice Hall series in computational mathematics.

[7]  Barbara M. Chapman,et al.  Extending HPF for Advanced Data-Parallel Applications , 1994, IEEE Parallel & Distributed Technology: Systems & Applications.

[8]  William Gropp,et al.  Domain decomposition on parallel computers , 1989, IMPACT Comput. Sci. Eng..

[9]  British Ornithologists,et al.  Bulletin of the , 1999 .

[10]  David E. Keyes,et al.  A Perspective On Data-Parallel Implicit Solvers For Mechanics , 1995 .

[11]  Guy L. Steele,et al.  The High Performance Fortran Handbook , 1993 .

[12]  Aaas News,et al.  Book Reviews , 1893, Buffalo Medical and Surgical Journal.

[13]  Piyush Mehrotra,et al.  Implementation of a Total Variation Diminishing Scheme for the Shock Tube Problem in High Performance Fortran , 1997, PPSC.

[14]  Rice UniversityCORPORATE,et al.  High performance Fortran language specification , 1993 .

[15]  David E. Keyes,et al.  A parallelized elliptic solver for reacting flows , 1987 .

[16]  Piyush Mehrotra,et al.  High Performance Fortran: History, Status and Future , 1998, Parallel Comput..

[17]  Y. Saad,et al.  GMRES: a generalized minimal residual algorithm for solving nonsymmetric linear systems , 1986 .

[18]  R. Dembo,et al.  INEXACT NEWTON METHODS , 1982 .

[19]  William Gropp,et al.  Parallel Newton-Krylov-Schwarz Algorithms for the Transonic Full Potential Equation , 1996, SIAM J. Sci. Comput..

[20]  Enrico Tronci.  1997 , 1997, Les 25 ans de l’OMC: Une rétrospective en photos.

[21]  William Gropp,et al.  Efficient Management of Parallelism in Object-Oriented Numerical Software Libraries , 1997, SciTools.