HP XC System 2.x Software - Manual (Page 125)

9
Using HP MLIB
The information in this section describes how to use HP MLIB Version 1.5 in the HP XC
environment on HP XC4000 and HP XC6000 clusters. These are discussed in separate sections
in this chapter.
9.1 Overview
HP MLIB is the mathematical library supported on the HP XC system. It is installed by default.
HP MLIB contains mathematical software and computational kernels for engineering and
scientific applications including the following:
• Linear equations
• Least squares
• Eigenvalue problems
• Singular value decomposition
• Vector and matrix computations
• Convolutions
• Fourier Transforms
HP MLIB V1.5 for XC6000 supports 64-bit addressing and 64-bit address libraries with all
integer values using 64 bits on XC6000 clusters.
HP MLIB for the HP XC4000 platform supports VECLIB, LAPACK, ScaLAPACK,
SuperLU_DIST, and MPI parallelism. This release does not support quad precision sparse
solver functionality on PGI/X86-64.
HP MLIB Version 1.5 is optimized for use on HP XC systems. HP MLIB Version 1.5 supports
64-bit addressing on XC6000 and XC4000 clusters.
9.1.1 Intel Compiler Notes
HP MLIB requires the Intel Fortran Compiler; the following provides information on Versions
8 and 7.1:
Version 8 Fortran Compiler
Command path name: /opt/mlib/intel_8.0/hpmpi_2.0
RPM: intel-ifort8-8.0-57.ia64.rpm
tar file: l_fc_pc_8[1].0.046.tar.gz
Version 7 Fortran Compiler
Command path name: /opt/mlib/intel_7.1/hpmpi_2.0
RPM: intel-efc7-7.1-41.ia64.rpm
tar file: l_fc_pc_7[1].1.040.tar
After installation, MLIB directory information is located in the /opt/mlib/README file.
Using HP MLIB 9-1