SuperLU

Refer to Part 3 of this manual for information specific to HP ScaLAPACK. To supplement the HP specific information provided in Part 3 of this document, refer to the standard ScaLAPACK Users’ Guide. You can access the latest edition of the ScaLAPACK Users’ Guide at the Netlib repository at the following URL:

http://www.netlib.org/scalapack/slug/index..html

SuperLU

This implementation provides the Distributed SuperLU library designed for distributed memory parallel computers. It was based on the public-domain SuperLU_DIST, which was developed at the Lawrence Berkeley National Lab and the University of California at Berkeley.

The library contains a set of subroutines to solve a sparse linear system. The library is implemented in ANSI C, using HP Message Passing Interface (MPI) for communication. The library includes routines to handle both real and complex matrices using double precision. The parallel routine names for the double-precision real version start with the letters “pd” (e.g. pdgstrf). The parallel routine names for the double-precision complex version start with letters “pz” (e.g. pzgstrf). Unlike other MLIB libraries, there is not a version of SuperLU that assumes all integers are 8 bytes in length.

The routines can be called directly from C applications. They may also be called from Fortran applications, however, “bridge” routines (in C) must be supplied. For details about creating bridge routines, please refer to Section 2.9.2 in the SuperLU User’s Guide available at:

http://www.nersc.gov/~xiaoye/SuperLU

SOLVERS

Solvers is a collection of direct sparse linear system solvers and graph partitioning routines. Symmetric systems can be solved using SMP parallelism and structurally-symmetric systems can be solved using out-of-core functionality. These routines have been optimized for use on Hewlett-Packard servers. Features are:

Preface xvii