Developer Reference for Intel® oneAPI Math Kernel Library for C

ID 766684
Date 11/07/2023
Public

A newer version of this document is available. Customers should click here to go to the newest version.

Document Table of Contents

mkl_set_num_stripes

Specifies the number of partitions along the leading dimension of the output matrix for parallel ?gemm functions.

Syntax

void mkl_set_num_stripes( int ns );

Include Files

  • mkl.h

Input Parameters

Name

Type

Description

ns

int

ns > 0 - Specifies the number of partitions to use.

ns= 0 - Instructs Intel® oneAPI Math Kernel Library (oneMKL) to use the default partitioning algorithm.

ns < 0 - Invalid value; ignored.

Description

This function enables you to specify the number of stripes, or partitions along the leading dimension of the output matrix, for parallel ?gemmfunctions. If this number is not set (default) or if it is set to zero, Intel® oneAPI Math Kernel Library (oneMKL)?gemm functions use the default partitioning algorithm. The specified number of partitions only applies to ?gemm functions.

The number specified is a hint, and Intel® oneAPI Math Kernel Library (oneMKL) may actually use a smaller number.

NOTE:

This function takes precedence over the MKL_NUM_STRIPES environment variable.

Example

#include "mkl.h"
…
mkl_set_num_stripes(4);
dgemm(...);	 // Intel MKL uses up to 4 stripes for dgemm