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
|
/*
* This file is a part of TiledArray.
* Copyright (C) 2013 Virginia Tech
*
* This program is free software: you can redistribute it and/or modify
* it under the terms of the GNU General Public License as published by
* the Free Software Foundation, either version 3 of the License, or
* (at your option) any later version.
*
* This program is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU General Public License for more details.
*
* You should have received a copy of the GNU General Public License
* along with this program. If not, see <http://www.gnu.org/licenses/>.
*
*/
#include <iostream>
#include <tiledarray.h>
template <typename it, typename tiletype>
void random_tile_task(it iter, tiletype tile){
std::size_t size = tile.size();
std::generate(tile.data(), tile.data()+size, []{return std::rand()%100;});
*iter = tile;
}
TiledArray::TArrayD
make_random_array(TiledArray::World &world, TiledArray::TiledRange &trange){
TiledArray::TArrayD array(world, trange);
typename TiledArray::TArrayD::iterator it = array.begin();
for(; it != array.end(); ++it){
typename TiledArray::TArrayD::value_type tile(
array.trange().make_tile_range(it.ordinal()));
world.taskq.add(&random_tile_task<decltype(it), decltype(tile)>, it, tile);
}
return array;
}
int main(int argc, char** argv) {
// Initialize runtime
TiledArray::World& world = TiledArray::initialize(argc, argv);
elem::Grid grid(elem::DefaultGrid().Comm());
// Get command line arguments
if(argc < 2) {
std::cout << "Usage: ta_dense matrix_size block_size [repetitions]\n";
return 0;
}
const long matrix_size = atol(argv[1]);
const long block_size = atol(argv[2]);
if (matrix_size <= 0) {
std::cerr << "Error: matrix size must be greater than zero.\n";
return 1;
}
if (block_size <= 0) {
std::cerr << "Error: block size must be greater than zero.\n";
return 1;
}
if((matrix_size % block_size) != 0ul) {
std::cerr << "Error: matrix size must be evenly divisible by block size.\n";
return 1;
}
const long repeat = (argc >= 4 ? atol(argv[3]) : 5);
if (repeat <= 0) {
std::cerr << "Error: number of repetitions must be greater than zero.\n";
return 1;
}
const std::size_t num_blocks = matrix_size / block_size;
const std::size_t block_count = num_blocks * num_blocks;
if(world.rank() == 0)
std::cout << "TiledArray: dense matrix multiply test...\n"
<< "Number of nodes = " << world.size()
<< "\nMatrix size = " << matrix_size << "x" << matrix_size
<< "\nBlock size = " << block_size << "x" << block_size
<< "\nMemory per matrix = " << double(matrix_size * matrix_size * sizeof(double)) / 1.0e9
<< " GB\nNumber of blocks = " << block_count
<< "\nAverage blocks/node = " << double(block_count) / double(world.size()) << "\n";
// Construct TiledRange
std::vector<unsigned int> blocking;
blocking.reserve(num_blocks + 1);
for(std::size_t i = 0; i <= matrix_size; i += block_size)
blocking.push_back(i);
std::vector<TiledArray::TiledRange1> blocking2(2,
TiledArray::TiledRange1(blocking.begin(), blocking.end()));
TiledArray::TiledRange
trange(blocking2.begin(), blocking2.end());
// Construct and initialize arrays
TiledArray::TArrayD a = make_random_array(world, trange);
TiledArray::TArrayD b = make_random_array(world, trange);
TiledArray::TArrayD c(world, trange);
if(world.rank() == 0 && matrix_size < 11){
std::cout << "a = \n" << a << std::endl;
std::cout << "b = \n" << b << std::endl;
}
// Start clock
world.gop.fence();
const double wall_time_start = madness::wall_time();
// Do matrix multiplication
for(int i = 0; i < repeat; ++i) {
c("m,n") = a("m,k") * b("k,n");
world.gop.fence();
if(world.rank() == 0)
std::cout << "Iteration " << i + 1 << "\n";
}
// Stop clock
const double wall_time_stop = madness::wall_time();
if(world.rank() == 0){
std::cout << "Average wall time = " << (wall_time_stop - wall_time_start) / double(repeat)
<< " sec\nAverage GFLOPS = " << double(repeat) * 2.0 * double(matrix_size *
matrix_size * matrix_size) / (wall_time_stop - wall_time_start) / 1.0e9 << "\n" << std::endl;
}
// Copying matrices to elemental
elem::DistMatrix<double> a_elem = array_to_elem(a,grid);
elem::DistMatrix<double> b_elem = array_to_elem(b,grid);
elem::mpi::Barrier(grid.Comm());
if(matrix_size < 11){
Print(a_elem, "a from elem");
Print(b_elem, "b from elem");
}
// Timed copy
const double wall_time_copy0 = madness::wall_time();
int j = 0;
while(j++ < repeat){
a_elem = array_to_elem(a,grid);
b_elem = array_to_elem(b,grid);
elem::mpi::Barrier(grid.Comm());
}
const double wall_time_copy1 = madness::wall_time();
// How long the copy took
if(world.rank() == 0){
std::cout << "Spent " <<
(wall_time_copy1 - wall_time_copy0)/(2.0 * double(repeat)) <<
" s for an array copy to elemental on average.\n" << std::endl;
}
// Make the data output array
elem::DistMatrix<double> c_elem(matrix_size, matrix_size, grid);
elem::Zero(c_elem);
elem::mpi::Barrier(grid.Comm());
// Do the multiply
const double wt_elem_start = madness::wall_time();
for(std::size_t i = 0; i < repeat; ++i){
elem::Gemm(elem::NORMAL, elem::NORMAL, 1., a_elem, b_elem, 0., c_elem);
elem::mpi::Barrier(grid.Comm());
if(grid.Rank() == 0){
std::cout << "Elem Iteration " << i + 1 << "\n";
}
}
const double wt_elem_end = madness::wall_time();
// Time elemental
if(world.rank() == 0){
std::cout << "Average Elemental wall time = " << (wt_elem_end - wt_elem_start) / double(repeat)
<< " sec\nAverage GFLOPS = " << double(repeat) * 2.0 * double(matrix_size *
matrix_size * matrix_size) / (wt_elem_end - wt_elem_start) / 1.0e9 << "\n";
}
// copy back to ta
int i = 0;
const double e_to_t_start = madness::wall_time();
while(i++ < repeat){
TiledArray::elem_to_array(c, c_elem);
elem::mpi::Barrier(grid.Comm());
}
const double e_to_t_end = madness::wall_time();
if(world.rank() == 0){
std::cout << "Copying to TA from Elemental took " << (e_to_t_end - e_to_t_start)/(double(repeat)) << " s on average." << std::endl;
}
TiledArray::finalize();
return 0;
}
|
