PETTT Advanced Computational Environments
- help@pettt-ace.com
- Name
- PETTT Advanced Computational Environments
ARL MSRC Linux Networx - MJM
- Architecture
- Intel Woodcrest
- BelongsTo.Organization
- ARL MSRC
- msrchelp@arl.army.mil
- Name
- ARL MSRC
- Telephone
- 1-800-275-1552
- Webpage
- http://www.arl.hpc.mil
- ClockSpeed
- 3.0 GHz
- DNS Name
- mjm-l[1-8].arl.hpc.mil
- DateOfInformation
- Fri May 30 17:10:19 PDT 2008
- Manufacturer
- Intel
- MemoryPerNode
- 16 GB
- Name
- ARL MSRC Linux Networx - MJM
- NetworkInterfaces
- DDR Infiniband
- OperatingSystem
- Linux
- PeakPerNode
- 48 Gigaflops
- ProcessorsPerNode
- 2 Dual Cores
- RecordMaintainedBy.Organization
- PETTT Advanced Computational Environments
- help@pettt-ace.com
- Name
- PETTT Advanced Computational Environments
- Webpage
- http://www.arl.hpc.mil/Systems/mjm.html
SuperLU
- Abstract
- SuperLU is a general purpose library for the direct solution of large, sparse, nonsymmetric systems of linear equations on high performance machines. The library is written in C and is callable from either C or Fortran. The library routines will perform an LU decomposition with partial pivoting and triangular system solvers through forward and back substitution. The LU factorization routines can handle non-square matrics but the triangular solves are performed only for square matrices. The matrix columns may be preordered (before factorization) either through library or user supplied routines. This preordering for sparsity is completely separate from the factorization. Working precision iterative refinement subroutines are provided for improved backward stability. Routines are also provided to equilibrate the system, estimate the condition number, calculate the relative backward error, and estimate error bounds for the refined solutions.
- DateOfInformation
- Thu Apr 15 16:40:08 EDT 2004
- Documentation
- http://crd.lbl.gov/~xiaoye/SuperLU/superlu_ug.ps.gz
- Domain
- Numerical Programs and Routines
- Linear algebra
- Name
- SuperLU
- TitleLine
- for direct solution of large, sparse linear systems
- Webpage
- http://crd.lbl.gov/~xiaoye/SuperLU/
There are three flavors of SuperLU: Sequential (Version 3.0), Multi-threaded (Version 1.0, shared-memory architecture, OpenMP, Pthreads) and Distributed (Version 2.0, distributed-memory architecture, MPI).
For example, the 64-bit Sequential version of SuperLU is in directory $PET_HOME/MATH/intel/lib" and is called "libsuperlu_SEQ64.a".
Header files are in subdirectories "include/SuperLU_SEQ", "include/SuperLU_DIST" and "include/SuperLU_MT".
To use SuperLU:
1) Need to have the appropriate "$PET_HOME/MATH/intel/include/..." directory in the include path, or specified with the "-I" compiler flag.
2) Need to have the appropriate SuperLU library in the library path, or specified with the "-l" and/or "-L" compiler flag.
3) Need to link with the C math library using "-lm" when building C programs, and need to link with the Fortran library using "-lfortran" when building Fortran programs.