-
Notifications
You must be signed in to change notification settings - Fork 15
/
pulp_matmul_fp32.h
239 lines (202 loc) · 7.05 KB
/
pulp_matmul_fp32.h
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
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
/*
* Copyright (C) 2021-2022 ETH Zurich and University of Bologna
*
* 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.
*/
/**
* Authors: Davide Nadalini, Leonardo Ravaglia
*/
/**
* Collection of different matrix multiply functions for different purposes
* Use pi_cl_team_fork(NUM_CORES, MM_NAME, &args) to parallelize.
*/
/**
* @brief Naive matrix multiply algorithm, performing C=A*B (C is N*M, A is N*K, B is K*M). Parallelizes on N.
* @param matMul_args pointer to a matMul_args structure (please refer to this to setup the args)
*/
void mm(
void * matMul_args
);
/**
* @brief Naive matrix multiply algorithm, performing C+=A*B (C is N*M, A is N*K, B is K*M). Parallelizes on N.
* @param matMul_args pointer to a matMul_args structure (please refer to this to setup the args)
*/
void mm_add(
void * matMul_args
);
/**
* @brief Naive matrix multiply algorithm, performing C=A*B (C is N*M, A is N*K, B is K*M). Parallelizes on M.
* @param matMul_args pointer to a matMul_args structure (please refer to this to setup the args)
*/
void mm_M(
void * matMul_args
);
/**
* Optimized versions
*/
// =====> PARALLELISM ON N <=====
/**
* @brief Standard matmul with unrolling factor of 2 in the inner loop on K, parallelizes on N.
* @param matMul_args pointer to a matMul_args structure (please refer to this to setup the args)
*/
void mm_u2(
void * matMul_args
);
/**
* @brief Standard matmul with unrolling, parallelizes on N. Unrolls 1 row of A, 2 columns of B.
* @param matMul_args pointer to a matMul_args structure (please refer to this to setup the args)
*/
void mm_unroll_1x2(
void * matMul_args
);
/**
* @brief Standard matmul with unrolling, parallelizes on N. Unrolls 1 row of A, 4 columns of B.
* @param matMul_args pointer to a matMul_args structure (please refer to this to setup the args)
*/
void mm_unroll_1x4(
void * matMul_args
);
/**
* @brief Standard matmul with unrolling, parallelizes on N. Unrolls 1 row of A, 8 columns of B.
* @param matMul_args pointer to a matMul_args structure (please refer to this to setup the args)
*/
void mm_unroll_1x8(
void * matMul_args
);
/**
* @brief Standard matmul with unrolling, parallelizes on N. Unrolls 2 row of A, 1 columns of B.
* @param matMul_args pointer to a matMul_args structure (please refer to this to setup the args)
*/
void mm_unroll_2x1(
void * matMul_args
);
/**
* @brief Standard matmul with unrolling, parallelizes on N. Unrolls 4 row of A, 1 columns of B.
* @param matMul_args pointer to a matMul_args structure (please refer to this to setup the args)
*/
void mm_unroll_4x1(
void * matMul_args
);
/**
* @brief Standard matmul with unrolling, parallelizes on N. Unrolls 8 row of A, 1 columns of B.
* @param matMul_args pointer to a matMul_args structure (please refer to this to setup the args)
*/
void mm_unroll_8x1(
void * matMul_args
);
/**
* @brief Standard matmul with unrolling, parallelizes on N. Unrolls 2 rows of A, 2 columns of B.
* @param matMul_args pointer to a matMul_args structure (please refer to this to setup the args)
*/
void mm_unroll_2x2(
void * matMul_args
);
/**
* @brief Standard matmul with unrolling, parallelizes on N. Unrolls 2 rows of A, 4 columns of B.
* @param matMul_args pointer to a matMul_args structure (please refer to this to setup the args)
*/
void mm_unroll_2x4(
void * matMul_args
);
/**
* @brief Standard matmul with unrolling, parallelizes on N. Unrolls 4 rows of A, 2 columns of B.
* @param matMul_args pointer to a matMul_args structure (please refer to this to setup the args)
*/
void mm_unroll_4x2(
void * matMul_args
);
/**
* @brief Standard matmul with unrolling, parallelizes on N. Unrolls 4 rows of A, 4 columns of B.
* @param matMul_args pointer to a matMul_args structure (please refer to this to setup the args)
*/
void mm_unroll_4x4(
void * matMul_args
);
// =====> PARALLELISM ON M <=====
/**
* @brief Naive matmul with unrolling factor of 2 in the inner loop on K, parallelizes on M.
* @param matMul_args pointer to a matMul_args structure (please refer to this to setup the args)
*/
void mm_M_u2(
void * matMul_args
);
/**
* @brief Standard matmul with unrolling, parallelizes on M. Unrolls 2 rows of A, 1 column of B.
* @param matMul_args pointer to a matMul_args structure (please refer to this to setup the args)
*/
void mm_M_unroll_2x1(
void * matMul_args
);
/**
* @brief Standard matmul with unrolling, parallelizes on M. Unrolls 4 rows of A, 1 column of B.
* @param matMul_args pointer to a matMul_args structure (please refer to this to setup the args)
*/
void mm_M_unroll_4x1(
void * matMul_args
);
/**
* @brief Standard matmul with unrolling, parallelizes on M. Unrolls 8 rows of A, 1 column of B.
* @param matMul_args pointer to a matMul_args structure (please refer to this to setup the args)
*/
void mm_M_unroll_8x1(
void * matMul_args
);
/**
* @brief Standard matmul with unrolling, parallelizes on M. Unrolls 1 row of A, 2 columns of B.
* @param matMul_args pointer to a matMul_args structure (please refer to this to setup the args)
*/
void mm_M_unroll_1x2(
void * matMul_args
);
/**
* @brief Standard matmul with unrolling, parallelizes on M. Unrolls 1 row of A, 4 columns of B.
* @param matMul_args pointer to a matMul_args structure (please refer to this to setup the args)
*/
void mm_M_unroll_1x4(
void * matMul_args
);
/**
* @brief Standard matmul with unrolling, parallelizes on M. Unrolls 1 row of A, 8 columns of B.
* @param matMul_args pointer to a matMul_args structure (please refer to this to setup the args)
*/
void mm_M_unroll_1x8(
void * matMul_args
);
/**
* @brief Standard matmul with unrolling, parallelizes on M. Unrolls 2 rows of A, 2 columns of B.
* @param matMul_args pointer to a matMul_args structure (please refer to this to setup the args)
*/
void mm_M_unroll_2x2(
void * matMul_args
);
/**
* @brief Standard matmul with unrolling, parallelizes on M. Unrolls 2 rows of A, 4 columns of B.
* @param matMul_args pointer to a matMul_args structure (please refer to this to setup the args)
*/
void mm_M_unroll_4x2(
void * matMul_args
);
/**
* @brief Standard matmul with unrolling, parallelizes on M. Unrolls 2 rows of A, 4 columns of B.
* @param matMul_args pointer to a matMul_args structure (please refer to this to setup the args)
*/
void mm_M_unroll_2x4(
void * matMul_args
);
/**
* @brief Standard matmul with unrolling, parallelizes on M. Unrolls 4 rows of A, 4 columns of B.
* @param matMul_args pointer to a matMul_args structure (please refer to this to setup the args)
*/
void mm_M_unroll_4x4(
void * matMul_args
);