Pakiet: libsuperlu-dist8 (8.1.2+dfsg1-1 i inne)

[ Pakiet źródłowy: superlu-dist  ]

Highly distributed solution of sparse linear equations

SuperLU is a general purpose library for the direct solution of large, sparse, nonsymmetric systems of linear equations. The library is written in C and is callable from either C or Fortran program. It uses MPI, OpenMP and CUDA to support various forms of parallelism. It supports both real and complex datatypes, both single and double precision, and 64-bit integer indexing. The library routines performs an LU decomposition with partial pivoting and triangular system solves through forward and back substitution. The LU factorization routines can handle non-square matrices 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.

SuperLU_DIST implements the algorithms for distributed memory, targetting highly parallel distributed memory hybrid systems. The numerical factorization routines are already implemented for hybrid systems with multiple GPUs. Further work will be needed to implement the other phases of the algorithms on the hybrid systems and to enhance strong scaling to extreme scale.

The main library is libsuperlu_dist.so but a fortran wrapper library is also provided as libsuperlu_dist_fortran.so

This package provides the superlu-dist and superlu_dist_fortran shared libraries.

Znaczniki: Rola: Biblioteka współdzielona

Inne pakiety związane z libsuperlu-dist8

• dep: libblas3

  Implementacje Basic Linear Algebra Reference, biblioteka współdzielona

  lub libblas.so.3

  pakiet wirtualny udostępniany przez libatlas3-base, libblas3, libblis4-openmp, libblis4-pthread, libblis4-serial, libopenblas0-openmp, libopenblas0-pthread, libopenblas0-serial

• dep: libc6 (>= 2.32)

  Biblioteka GNU C: biblioteki współdzielone
  również pakiet wirtualny udostępniany przez libc6-udeb

• dep: libcombblas1.16.0 (>= 1.6.2)

  an extensible parallel graph library for graph analytics

• dep: libgcc-s1 (>= 3.0) [nie armel, armhf, i386, mipsel]

  Biblioteka wspomagająca GCC

  dep: libgcc-s1 (>= 3.5) [armel, armhf]

  dep: libgcc-s1 (>= 4.2) [i386, mipsel]

• dep: libgomp1 (>= 6)

  Biblioteka wspierająca GCC OpenMP (GOMP)

• dep: liblapack3

  Library of linear algebra routines 3 - shared version

  lub liblapack.so.3

  pakiet wirtualny udostępniany przez libatlas3-base, liblapack3, libopenblas0-openmp, libopenblas0-pthread, libopenblas0-serial

• dep: libmetis5 (>= 5.1.0.dfsg)

  Serial Graph Partitioning and Fill-reducing Matrix Ordering

• dep: libopenmpi3 (>= 4.1.4)

  high performance message passing library -- shared library

• dep: libptscotch-7.0 (>= 7.0.3)

  MPI programs and libraries for graph, mesh and hypergraph partitioning

• dep: libstdc++6 (>= 11)

  Standardowa biblioteka GNU C++, wersja 3

Pobieranie libsuperlu-dist8

Pobierz dla wszystkich dostępnych architektur

Architektura	Wersja	Rozmiar pakietu	Rozmiar po instalacji	Pliki
amd64	8.1.2+dfsg1-1+b1	581,0 KiB	1 893,0 KiB	[lista plików]
arm64	8.1.2+dfsg1-1+b1	499,5 KiB	1 761,0 KiB	[lista plików]
armel	8.1.2+dfsg1-1+b1	494,4 KiB	1 615,0 KiB	[lista plików]
armhf	8.1.2+dfsg1-1+b1	466,7 KiB	1 111,0 KiB	[lista plików]
i386	8.1.2+dfsg1-1+b1	559,3 KiB	2 007,0 KiB	[lista plików]
mips64el	8.1.2+dfsg1-1+b1	450,6 KiB	2 033,0 KiB	[lista plików]
mipsel	8.1.2+dfsg1-1+b1	476,2 KiB	2 002,0 KiB	[lista plików]
ppc64el	8.1.2+dfsg1-1+b1	605,7 KiB	2 273,0 KiB	[lista plików]
s390x	8.1.2+dfsg1-1+b1	524,5 KiB	2 021,0 KiB	[lista plików]