tatami_mult::MultiplyDenseRowWithDenseColumnMatrixToColumnOutputOptions Struct Reference

Options for multiply_dense_row_with_dense_column_matrix_to_column_output(). More...

#include <column_to_column.hpp>

Public Attributes
int	num_threads = 1
int	primary_block_size = 16
int	secondary_block_size = 64

Detailed Description

Options for multiply_dense_row_with_dense_column_matrix_to_column_output().

Member Data Documentation

num_threads

int tatami_mult::MultiplyDenseRowWithDenseColumnMatrixToColumnOutputOptions::num_threads = 1

Number of threads to use. Different numbers of threads will not change the results.

primary_block_size

int tatami_mult::MultiplyDenseRowWithDenseColumnMatrixToColumnOutputOptions::primary_block_size = 16

Primary block size, i.e., the number of LHS rows to be loaded at once. This is also used to define the number of RHS columns in each block. See the \(B\) parameter in the Blocking for dense matrices section for more details.

secondary_block_size

int tatami_mult::MultiplyDenseRowWithDenseColumnMatrixToColumnOutputOptions::secondary_block_size = 64

Secondary block size, i.e., the number of LHS columns to be processed in each block. See the \(C\) parameter in the Blocking for dense matrices section for more details. Different secondary block sizes may slightly change the results due to differences in floating-point round-off error.

---

The documentation for this struct was generated from the following file:

• tatami_mult/dense_matrix/dense_row/column_to_column.hpp