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
|
/******************************************************************************
* Copyright (c) Intel Corporation - All rights reserved. *
* This file is part of the LIBXSMM library. *
* *
* For information on the license, see the LICENSE file. *
* Further information: https://github.com/hfp/libxsmm/ *
* SPDX-License-Identifier: BSD-3-Clause *
******************************************************************************/
/* Alexander Heinecke (Intel Corp.)
******************************************************************************/
#include <stdio.h>
#include <stdlib.h>
#include <math.h>
#include <libxsmm.h>
int main(int argc, char* argv[]) {
unsigned int N = ( argc > 1 ) ? atoi(argv[1]) : 64;
unsigned int C = ( argc > 2 ) ? atoi(argv[2]) : 512;
unsigned int K = ( argc > 3 ) ? atoi(argv[3]) : 32;
unsigned int nb = ( argc > 4 ) ? atoi(argv[4]) : 16;
double sparse_frac = ( argc > 5 ) ? atof(argv[5]) : 0.90;
unsigned int REPS = ( argc > 6 ) ? atoi(argv[6]) : 1;
const libxsmm_gemm_prefetch_type prefetch = LIBXSMM_GEMM_PREFETCH_NONE;
const int flags = LIBXSMM_GEMM_FLAGS('N', 'N');
const float alpha = 1, beta = 1;
unsigned int* l_colptr = NULL;
unsigned int* l_rowidx = NULL;
unsigned int* l_rowptr = NULL;
unsigned int* l_colidx = NULL;
float* l_b_de = (float*)libxsmm_aligned_malloc(sizeof(float) * C * K, 64);
float* l_b_sp_csc = NULL;
float* l_b_sp_csr = NULL;
float* l_a = (float*)libxsmm_aligned_malloc(sizeof(float) * N * C, 64);
float* l_c_gold = (float*)libxsmm_aligned_malloc(sizeof(float) * N * K, 64);
float* l_c_asm_csc = (float*)libxsmm_aligned_malloc(sizeof(float) * N * K, 64);
float* l_c_asm_csr = (float*)libxsmm_aligned_malloc(sizeof(float) * N * K, 64);
float l_max_error = 0.0;
unsigned int l_k, l_n;
unsigned int l_i, l_j, l_jj;
LIBXSMM_VLA_DECL(2, float, l_p_b_de, l_b_de, C);
LIBXSMM_VLA_DECL(3, float, l_p_a, l_a, C, nb);
LIBXSMM_VLA_DECL(3, float, l_p_c_asm_csc, l_c_asm_csc, K, nb);
LIBXSMM_VLA_DECL(3, float, l_p_c_asm_csr, l_c_asm_csr, K, nb);
LIBXSMM_VLA_DECL(3, float, l_p_c_gold, l_c_gold, K, nb);
libxsmm_descriptor_blob l_xgemm_blob;
const libxsmm_gemm_descriptor* l_xgemm_desc = 0;
LIBXSMM_MMFUNCTION_TYPE(float) mykernel_csc = NULL;
LIBXSMM_MMFUNCTION_TYPE(float) mykernel_csr = NULL;
unsigned long long l_start, l_end;
double l_total;
unsigned int NB;
unsigned int nnz = 0;
if (argc != 7 && argc != 1) {
fprintf( stderr, "arguments failure\n" );
return -1;
}
NB = N / nb;
/* touch A */
for ( l_i = 0; l_i < NB; l_i++) {
for ( l_j = 0; l_j < C; l_j++) {
for ( l_k = 0; l_k < nb; l_k++ ) {
LIBXSMM_VLA_ACCESS(3, l_p_a, l_i, l_j, l_k, C, nb) = (float)libxsmm_rng_f64();
}
}
}
/* touch dense B */
for ( l_i = 0; l_i < K; l_i++ ) {
for ( l_j = 0; l_j < C; l_j++ ) {
float tmp = (float)libxsmm_rng_f64();
if ( tmp < sparse_frac ) {
tmp = 0;
} else {
nnz++;
}
LIBXSMM_VLA_ACCESS(2, l_p_b_de, l_i, l_j, C) = tmp;
}
}
printf("we just generated a %i x %i matrix with %i NZ entries\n", K, C, nnz);
/* touch C */
for ( l_i = 0; l_i < NB; l_i++) {
for ( l_j = 0; l_j < K; l_j++) {
for ( l_k = 0; l_k < nb; l_k++ ) {
LIBXSMM_VLA_ACCESS(3, l_p_c_gold, l_i, l_j, l_k, K, nb) = 0.f;
LIBXSMM_VLA_ACCESS(3, l_p_c_asm_csc, l_i, l_j, l_k, K, nb) = 0.f;
LIBXSMM_VLA_ACCESS(3, l_p_c_asm_csr, l_i, l_j, l_k, K, nb) = 0.f;
}
}
}
/* create B, csc */
l_colptr = (unsigned int*) libxsmm_aligned_malloc( (K+1)*sizeof(unsigned int), 64 );
l_rowidx = (unsigned int*) libxsmm_aligned_malloc( nnz*sizeof(unsigned int), 64 );
l_b_sp_csc = (float* ) libxsmm_aligned_malloc( nnz*sizeof(float), 64 );
l_k = 0;
l_colptr[K] = nnz;
for ( l_i = 0; l_i < K; l_i++ ) {
l_colptr[l_i] = l_k;
for ( l_j = 0; l_j < C; l_j++ ) {
if ( LIBXSMM_VLA_ACCESS(2, l_p_b_de, l_i, l_j, C) != 0.0 ) {
l_rowidx[l_k] = l_j;
l_b_sp_csc[l_k] = LIBXSMM_VLA_ACCESS(2, l_p_b_de, l_i, l_j, C);
l_k++;
}
}
}
/* create B, csr */
l_rowptr = (unsigned int*) libxsmm_aligned_malloc( (C+1)*sizeof(unsigned int), 64 );
l_colidx = (unsigned int*) libxsmm_aligned_malloc( nnz*sizeof(unsigned int), 64 );
l_b_sp_csr = (float* ) libxsmm_aligned_malloc( nnz*sizeof(float), 64 );
l_k = 0;
l_rowptr[C] = nnz;
for ( l_j = 0; l_j < C; l_j++ ) {
l_rowptr[l_j] = l_k;
for ( l_i = 0; l_i < K; l_i++ ) {
if ( LIBXSMM_VLA_ACCESS(2, l_p_b_de, l_i, l_j, C) != 0.0 ) {
l_colidx[l_k] = l_i;
l_b_sp_csr[l_k] = LIBXSMM_VLA_ACCESS(2, l_p_b_de, l_i, l_j, C);
l_k++;
}
}
}
/* dense routine */
l_start = libxsmm_timer_tick();
#if 1
for ( l_n = 0; l_n < REPS; l_n++) {
for ( l_i = 0; l_i < NB; l_i++) {
for ( l_j = 0; l_j < K; l_j++) {
for ( l_jj = 0; l_jj < C; l_jj++) {
LIBXSMM_PRAGMA_SIMD
for (l_k = 0; l_k < nb; l_k++) {
LIBXSMM_VLA_ACCESS(3, l_p_c_gold, l_i, l_j, l_k, K, nb)
+= LIBXSMM_VLA_ACCESS(3, l_p_a, l_i, l_jj, l_k, C, nb)
* l_b_de[(l_j*C)+l_jj];
}
}
}
}
}
#endif
l_end = libxsmm_timer_tick();
l_total = libxsmm_timer_duration(l_start, l_end);
printf("%fs for dense\n", l_total);
printf("%f GFLOPS for dense\n", ((double)((double)REPS * (double)N * (double)C * (double)K) * 2.0) / (l_total * 1.0e9));
l_xgemm_desc = libxsmm_gemm_descriptor_dinit(&l_xgemm_blob, LIBXSMM_GEMM_PRECISION(float),
NB, K, C, C, 0, K, alpha, beta, flags, prefetch);
/* sparse routine */
mykernel_csc = libxsmm_create_xcsc_soa(l_xgemm_desc, l_colptr, l_rowidx, (const void*)l_b_sp_csc, nb).smm;
mykernel_csr = libxsmm_create_xcsr_soa(l_xgemm_desc, l_rowptr, l_colidx, (const void*)l_b_sp_csr, nb).smm;
l_start = libxsmm_timer_tick();
for ( l_n = 0; l_n < REPS; l_n++) {
mykernel_csc( l_a, l_b_sp_csc, l_c_asm_csc );
}
l_end = libxsmm_timer_tick();
l_total = libxsmm_timer_duration(l_start, l_end);
printf("%fs for sparse (asm, csc)\n", l_total);
printf("%f GFLOPS for sparse (asm, csc)\n", ((double)((double)REPS * (double)N * (double)C * (double)K) * 2.0) / (l_total * 1.0e9));
l_start = libxsmm_timer_tick();
for ( l_n = 0; l_n < REPS; l_n++) {
mykernel_csr( l_a, l_b_sp_csr, l_c_asm_csr );
}
l_end = libxsmm_timer_tick();
l_total = libxsmm_timer_duration(l_start, l_end);
printf("%fs for sparse (asm, csr)\n", l_total);
printf("%f GFLOPS for sparse (asm, csr)\n", ((double)((double)REPS * (double)N * (double)C * (double)K) * 2.0) / (l_total * 1.0e9));
/* check for errors */
l_max_error = 0.f;
for ( l_i = 0; l_i < NB; l_i++) {
for ( l_j = 0; l_j < K; l_j++) {
for ( l_k = 0; l_k < nb; l_k++ ) {
if (fabs( LIBXSMM_VLA_ACCESS(3, l_p_c_gold, l_i, l_j, l_k, K, nb)
- LIBXSMM_VLA_ACCESS(3, l_p_c_asm_csc, l_i, l_j, l_k, K, nb) ) > l_max_error ) {
l_max_error = (float)fabs( LIBXSMM_VLA_ACCESS(3, l_p_c_gold, l_i, l_j, l_k, K, nb)
-LIBXSMM_VLA_ACCESS(3, l_p_c_asm_csc, l_i, l_j, l_k, K, nb) );
}
}
}
}
printf("max error (csc): %f\n", l_max_error);
l_max_error = 0.f;
for ( l_i = 0; l_i < NB; l_i++) {
for ( l_j = 0; l_j < K; l_j++) {
for ( l_k = 0; l_k < nb; l_k++ ) {
if (fabs( LIBXSMM_VLA_ACCESS(3, l_p_c_gold, l_i, l_j, l_k, K, nb)
- LIBXSMM_VLA_ACCESS(3, l_p_c_asm_csr, l_i, l_j, l_k, K, nb) ) > l_max_error ) {
l_max_error = (float)fabs( LIBXSMM_VLA_ACCESS(3, l_p_c_gold, l_i, l_j, l_k, K, nb)
-LIBXSMM_VLA_ACCESS(3, l_p_c_asm_csr, l_i, l_j, l_k, K, nb) );
}
}
}
}
printf("max error (csr): %f\n", l_max_error);
/* free */
libxsmm_free( l_b_de );
libxsmm_free( l_a );
libxsmm_free( l_c_gold );
libxsmm_free( l_c_asm_csc );
libxsmm_free( l_c_asm_csr );
libxsmm_free( l_b_sp_csc );
libxsmm_free( l_colptr );
libxsmm_free( l_rowidx );
libxsmm_free( l_b_sp_csr );
libxsmm_free( l_rowptr );
libxsmm_free( l_colidx );
return 0;
}
|
