CodeVault: Sparse Linear Algebra ================ # Overview In this Sparse Linear Algebra folder, there are sample codes for beginners, as well as advanced codes including important computational kernels. # Contributors & Maintainers - Cevdet Aykanat (aykanat@cs.bilkent.edu.tr) - Kadir Akbudak (kadir.cs@gmail.com) - Reha Oguz Selvitopi(reha@cs.bilkent.edu.tr) # Contents - spmv: Multiplication of a sparse matrix with a dense vector. - mkl_shmem: Using MKL's routine mkl_dcsrmv() on a multicore processor - spgemm: Multiplication of two sparse matrices. - mkl_shmem: Using MKL's routine mkl_dcsrmultcsr() on a multicore processor - mkl_xphi: Using MKL's routine mkl_dcsrmultcsr() via offloading to a Xeon Phi coprocessor These two codes also contain schoolbook implementation of an SpGEMM algorithm that uses row-by-row formulation [1]. - Krylov Subspace Methods - Linear system solution in parallel - 2D Laplacian (2D mesh) - Repeatedly solving two linear systems - Same preconditioner - Two different matrices with the same nonzero pattern - Solving multiple linear systems - Same cofficient matrix - Different right-hand-side vectors REFERENCES: [1] Fred G. Gustavson. 1978. Two Fast Algorithms for Sparse Matrices: Multiplication and Permuted Transposition. ACM Trans. Math. Softw. 4, 3 (September 1978), 250-269. DOI=http://dx.doi.org/10.1145/355791.355796