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
|
// Copyright Yahoo. Licensed under the terms of the Apache 2.0 license. See LICENSE in the project root.
#include <vespa/vespalib/hwaccelrated/iaccelrated.h>
#include <vespa/vespalib/stllike/string.h>
#include <vespa/vespalib/stllike/hash_map.h>
#include <iostream>
#include <vector>
#include <thread>
#include <functional>
using namespace vespalib;
using vespalib::hwaccelrated::IAccelrated;
class Benchmark {
public:
virtual ~Benchmark() { }
virtual void compute(size_t docId) const = 0;
};
void
runThread(size_t count, size_t docs, const Benchmark * benchmark, size_t stride)
{
for (size_t i(0); i < count; i++) {
for (size_t docId(0); docId < docs; docId++) {
benchmark->compute((docId*stride)%docs);
}
}
}
void
runBenchmark(size_t numThreads, size_t count, size_t docs, const Benchmark & benchmark, size_t stride) {
if (numThreads > 1) {
std::vector<std::thread> threads;
threads.reserve(numThreads);
for (size_t i(0); i < numThreads; i++) {
threads.emplace_back(runThread, count, docs, &benchmark, stride);
}
for (auto & thread : threads) {
thread.join();
}
} else {
runThread(count, docs, &benchmark, stride);
}
}
template <typename T>
class FullBenchmark : public Benchmark
{
public:
FullBenchmark(size_t numDocs, size_t numValue);
~FullBenchmark();
void compute(size_t docId) const override {
_dp.dotProduct(&_query[0], &_values[docId * _query.size()], _query.size());
}
private:
std::vector<T> _values;
std::vector<T> _query;
const IAccelrated & _dp;
};
template <typename T>
FullBenchmark<T>::FullBenchmark(size_t numDocs, size_t numValues)
: _values(numDocs*numValues),
_query(numValues),
_dp(IAccelrated::getAccelerator())
{
for (size_t i(0); i < numDocs; i++) {
for (size_t j(0); j < numValues; j++) {
_values[i*numValues + j] = j;
}
}
for (size_t j(0); j < numValues; j++) {
_query[j] = j;
}
}
template <typename T>
FullBenchmark<T>::~FullBenchmark() = default;
class SparseBenchmark : public Benchmark
{
public:
SparseBenchmark(size_t numDocs, size_t numValues, size_t numQueryValues);
~SparseBenchmark();
protected:
struct P {
P(uint32_t key=0, int32_t value=0) noexcept :
_key(key),
_value(value)
{ }
uint32_t _key;
int32_t _value;
};
size_t _numValues;
std::vector<P> _values;
};
SparseBenchmark::SparseBenchmark(size_t numDocs, size_t numValues, size_t numQueryValues)
: _numValues(numValues),
_values(numDocs*numValues)
{
for (size_t i(0); i < numDocs; i++) {
for (size_t j(0); j < numValues; j++) {
size_t k(numValues < numQueryValues ? (j*numQueryValues)/numValues : j);
_values[i*numValues + j] = P(k, k);
}
}
}
SparseBenchmark::~SparseBenchmark() = default;
std::function<void(int64_t)> use_sum = [](int64_t) noexcept { };
class UnorderedSparseBenchmark : public SparseBenchmark
{
private:
using map = hash_map<uint32_t, int32_t>;
public:
UnorderedSparseBenchmark(size_t numDocs, size_t numValues, size_t numQueryValues);
~UnorderedSparseBenchmark();
private:
void compute(size_t docId) const override {
int64_t sum(0);
size_t offset(docId*_numValues);
const auto e(_query.end());
for (size_t i(0); i < _numValues; i++) {
auto it = _query.find(_values[offset + i]._key);
if (it != e) {
sum += static_cast<int64_t>(_values[offset + i]._value) * it->second;
}
}
use_sum(sum);
}
map _query;
};
UnorderedSparseBenchmark::UnorderedSparseBenchmark(size_t numDocs, size_t numValues, size_t numQueryValues)
: SparseBenchmark(numDocs, numValues, numQueryValues)
{
for (size_t j(0); j < numQueryValues; j++) {
_query[j] = j;
}
}
UnorderedSparseBenchmark::~UnorderedSparseBenchmark() = default;
class OrderedSparseBenchmark : public SparseBenchmark
{
private:
public:
OrderedSparseBenchmark(size_t numDocs, size_t numValues, size_t numQueryValues);
~OrderedSparseBenchmark();
private:
void compute(size_t docId) const override {
int64_t sum(0);
size_t offset(docId*_numValues);
for (size_t a(0), b(0); a < _query.size() && b < _numValues; b++) {
for (; a < _query.size() && (_query[a]._key <= _values[offset + b]._key); a++);
if (_query[a]._key == _values[offset + b]._key) {
sum += static_cast<int64_t>(_values[offset + b]._value) * _query[a]._value;
}
}
use_sum(sum);
}
std::vector<P> _query;
};
OrderedSparseBenchmark::OrderedSparseBenchmark(size_t numDocs, size_t numValues, size_t numQueryValues)
: SparseBenchmark(numDocs, numValues, numQueryValues),
_query(numQueryValues)
{
for (size_t j(0); j < numQueryValues; j++) {
size_t k(numValues > numQueryValues ? j*numValues/numQueryValues : j);
_query[j] = P(k, k);
}
}
OrderedSparseBenchmark::~OrderedSparseBenchmark() = default;
int main(int argc, char *argv[])
{
size_t numThreads(1);
size_t numDocs(1);
size_t numValues(1000);
size_t numQueries(1000000);
size_t stride(1);
string type("full");
if ( argc > 1) {
type = argv[1];
}
if ( argc > 2) {
numThreads = strtoul(argv[2], nullptr, 0);
}
if ( argc > 3) {
numQueries = strtoul(argv[3], nullptr, 0);
}
if ( argc > 4) {
numDocs = strtoul(argv[4], nullptr, 0);
}
if ( argc > 5) {
numValues = strtoul(argv[5], nullptr, 0);
}
if (argc > 6) {
stride = strtoul(argv[6], nullptr, 0);
}
size_t numQueryValues = ( argc > 7) ? strtoul(argv[7], nullptr, 0) : numValues;
std::cout << "type = " << type << std::endl;
std::cout << "numQueries = " << numQueries << std::endl;
std::cout << "numDocs = " << numDocs << std::endl;
std::cout << "numValues = " << numValues << std::endl;
std::cout << "numQueryValues = " << numQueryValues << std::endl;
std::cout << "stride =" << stride << std::endl;
if (type == "full") {
FullBenchmark<float> bm(numDocs, numValues);
runBenchmark(numThreads, numQueries, numDocs, bm, stride);
} else if (type == "sparse-ordered") {
OrderedSparseBenchmark bm(numDocs, numValues, numQueryValues);
runBenchmark(numThreads, numQueries, numDocs, bm, stride);
} else if (type == "sparse-unordered") {
UnorderedSparseBenchmark bm(numDocs, numValues, numQueryValues);
runBenchmark(numThreads, numQueries, numDocs, bm, stride);
} else {
std::cerr << "type '" << type << "' is unknown." << std::endl;
}
return 0;
}
|
