1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
|
/*******************************************************************************
* Copyright 2020-2025 Intel Corporation
*
* Licensed under the Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*******************************************************************************/
/// @example matmul.cpp
/// > Annotated version: @ref matmul_example_cpp
///
/// @page matmul_example_cpp_short
///
/// This C++ API example demonstrates how to create and execute a
/// [MatMul](@ref dev_guide_matmul) primitive.
///
/// Key optimizations included in this example:
/// - Primitive attributes with fused post-ops.
///
/// @page matmul_example_cpp Matmul Primitive Example
/// @copydetails matmul_example_cpp_short
///
/// @include matmul.cpp
#include <algorithm>
#include <cmath>
#include <iostream>
#include <string>
#include <vector>
#include "example_utils.hpp"
#include "oneapi/dnnl/dnnl.hpp"
using namespace dnnl;
void matmul_example(dnnl::engine::kind engine_kind) {
// Create execution dnnl::engine.
dnnl::engine engine(engine_kind, 0);
// Create dnnl::stream.
dnnl::stream engine_stream(engine);
// Tensor dimensions.
const memory::dim MB = 3, // batch size
M = 128, K = 256, N = 512;
// Source (src), weights, bias, and destination (dst) tensors dimensions.
memory::dims src_dims = {MB, M, K};
memory::dims weights_dims = {MB, K, N};
memory::dims bias_dims = {1, 1, N};
memory::dims dst_dims = {MB, M, N};
// Allocate buffers.
std::vector<float> src_data(product(src_dims));
std::vector<float> weights_data(product(weights_dims));
std::vector<float> bias_data(product(bias_dims));
std::vector<float> dst_data(product(dst_dims));
// Initialize src, weights, bias.
std::generate(src_data.begin(), src_data.end(), []() {
static int i = 0;
return std::cos(i++ / 10.f);
});
std::generate(weights_data.begin(), weights_data.end(), []() {
static int i = 0;
return std::sin(i++ * 2.f);
});
std::generate(bias_data.begin(), bias_data.end(), []() {
static int i = 0;
return std::tanh(float(i++));
});
// Create memory descriptors and memory objects for src, weights, bias, and
// dst.
auto src_md = memory::desc(
src_dims, memory::data_type::f32, memory::format_tag::abc);
auto weights_md = memory::desc(
weights_dims, memory::data_type::f32, memory::format_tag::abc);
auto bias_md = memory::desc(
bias_dims, memory::data_type::f32, memory::format_tag::abc);
auto dst_md = memory::desc(
dst_dims, memory::data_type::f32, memory::format_tag::abc);
auto src_mem = memory(src_md, engine);
auto weights_mem = memory(weights_md, engine);
auto bias_mem = memory(bias_md, engine);
auto dst_mem = memory(dst_md, engine);
// Write data to memory object's handles.
write_to_dnnl_memory(src_data.data(), src_mem);
write_to_dnnl_memory(weights_data.data(), weights_mem);
write_to_dnnl_memory(bias_data.data(), bias_mem);
// Create primitive post-ops (ReLU).
const float alpha = 0.f;
const float beta = 0.f;
post_ops matmul_ops;
matmul_ops.append_eltwise(algorithm::eltwise_relu, alpha, beta);
primitive_attr matmul_attr;
matmul_attr.set_post_ops(matmul_ops);
// Create primitive descriptor.
auto matmul_pd = matmul::primitive_desc(
engine, src_md, weights_md, bias_md, dst_md, matmul_attr);
// Create the primitive.
auto matmul_prim = matmul(matmul_pd);
// Primitive arguments.
std::unordered_map<int, memory> matmul_args;
matmul_args.insert({DNNL_ARG_SRC, src_mem});
matmul_args.insert({DNNL_ARG_WEIGHTS, weights_mem});
matmul_args.insert({DNNL_ARG_BIAS, bias_mem});
matmul_args.insert({DNNL_ARG_DST, dst_mem});
// Primitive execution: matrix multiplication with ReLU.
matmul_prim.execute(engine_stream, matmul_args);
// Wait for the computation to finalize.
engine_stream.wait();
// Read data from memory object's handle.
read_from_dnnl_memory(dst_data.data(), dst_mem);
}
int main(int argc, char **argv) {
return handle_example_errors(matmul_example, parse_engine_kind(argc, argv));
}
|
