Intel® oneAPI Deep Neural Network Library Developer Guide and Reference

Developer Guide and Reference

  • 2022.1
  • 04/11/2022
  • Public Content
Contents

struct dnnl::matmul::desc

Overview

Descriptor for a matmul primitive. More…
#include <dnnl.hpp>

struct desc
{
    // fields

    dnnl_matmul_desc_t data;

    // construction

    desc(
        const memory::desc& src_desc,
        const memory::desc& weights_desc,
        const memory::desc& dst_desc
        );

    desc(
        const memory::desc& src_desc,
        const memory::desc& weights_desc,
        const memory::desc& bias_desc,
        const memory::desc& dst_desc
        );
};

Detailed Documentation

Descriptor for a matmul primitive.
Construction
desc(
    const memory::desc& src_desc,
    const memory::desc& weights_desc,
    const memory::desc& dst_desc
    )
Constructs a descriptor for a matmul primitive.
Parameters:
src_desc
Memory descriptor for source (matrix A).
weights_desc
Memory descriptor for weights (matrix B).
dst_desc
Memory descriptor for destination (matrix C).
 
desc(
    const memory::desc& src_desc,
    const memory::desc& weights_desc,
    const memory::desc& bias_desc,
    const memory::desc& dst_desc
    )
Constructs a descriptor for a matmul primitive.
Parameters:
src_desc
Memory descriptor for source (matrix A).
weights_desc
Memory descriptor for weights (matrix B).
dst_desc
Memory descriptor for destination (matrix C).
bias_desc
Memory descriptor for bias.

Product and Performance Information

1

Performance varies by use, configuration and other factors. Learn more at www.Intel.com/PerformanceIndex.