MATAIJHIPSPARSE#

A matrix type to be used for sparse matrices; it is as same as MATMPIAIJHIPSPARSE. A matrix type type whose data resides on Nvidia GPUs. These matrices can be in either CSR, ELL, or Hybrid format. All matrix calculations are performed on AMD GPUs using the HIPSPARSE library.

This matrix type is identical to MATSEQAIJHIPSPARSE when constructed with a single process communicator, and MATMPIAIJHIPSPARSE otherwise. As a result, for single process communicators, MatSeqAIJSetPreallocation is supported, and similarly MatMPIAIJSetPreallocation is supported for communicators controlling multiple processes. It is recommended that you call both of the above preallocation routines for simplicity.

Options Database Keys#

  • -mat_type mpiaijhipsparse - sets the matrix type to “mpiaijhipsparse” during a call to MatSetFromOptions()

  • -mat_hipsparse_storage_format csr - sets the storage format of diagonal and off-diagonal matrices during a call to MatSetFromOptions(). Other options include ell (ellpack) or hyb (hybrid).

  • -mat_hipsparse_mult_diag_storage_format csr - sets the storage format of diagonal matrix during a call to MatSetFromOptions(). Other options include ell (ellpack) or hyb (hybrid).

  • -mat_hipsparse_mult_offdiag_storage_format csr - sets the storage format of off-diagonal matrix during a call to MatSetFromOptions(). Other options include ell (ellpack) or hyb (hybrid).

.seealso: MatCreateAIJHIPSPARSE(), MATSEQAIJHIPSPARSE, MATMPIAIJHIPSPARSE, MatCreateSeqAIJHIPSPARSE(), MatHIPSPARSESetFormat(), MatHIPSPARSEStorageFormat, MatHIPSPARSEFormatOperation

Level#

beginner

Location#

src/mat/impls/aij/mpi/mpihipsparse/mpiaijhipsparse.hip.cpp


Edit on GitLab

Index of all Mat routines
Table of Contents for all manual pages
Index of all manual pages