Repository navigation
Expand file tree
/
Copy pathBenchmarkTaskflow.html
More file actions
263 lines (261 loc) · 15.8 KB
/
Copy pathBenchmarkTaskflow.html
File metadata and controls
263 lines (261 loc) · 15.8 KB
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
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
<!-- HTML header for doxygen 1.13.1-->
<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "https://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
<html xmlns="http://www.w3.org/1999/xhtml" lang="en-US">
<head>
<meta http-equiv="Content-Type" content="text/xhtml;charset=UTF-8"/>
<meta http-equiv="X-UA-Compatible" content="IE=11"/>
<meta name="generator" content="Doxygen 1.13.1"/>
<meta name="viewport" content="width=device-width, initial-scale=1"/>
<title>Taskflow: A General-purpose Task-parallel Programming System: Benchmark Taskflow</title>
<link href="tabs.css" rel="stylesheet" type="text/css"/>
<script type="text/javascript" src="jquery.js"></script>
<script type="text/javascript" src="dynsections.js"></script>
<script type="text/javascript" src="clipboard.js"></script>
<link href="navtree.css" rel="stylesheet" type="text/css"/>
<script type="text/javascript" src="navtreedata.js"></script>
<script type="text/javascript" src="navtree.js"></script>
<script type="text/javascript" src="resize.js"></script>
<script type="text/javascript" src="cookie.js"></script>
<link href="search/search.css" rel="stylesheet" type="text/css"/>
<script type="text/javascript" src="search/searchdata.js"></script>
<script type="text/javascript" src="search/search.js"></script>
<script type="text/javascript" src="darkmode_toggle.js"></script>
<link href="doxygen.css" rel="stylesheet" type="text/css" />
<link href="custom.css" rel="stylesheet" type="text/css"/>
</head>
<body>
<div id="top"><!-- do not remove this div, it is closed by doxygen! -->
<div id="titlearea">
<table cellspacing="0" cellpadding="0">
<tbody>
<tr id="projectrow">
<td id="projectlogo"><img alt="Logo" src="taskflow_logo.png"/></td>
<td id="projectalign">
<div id="projectname"><a href="https://github.com/taskflow/taskflow" style="color:inherit; text-decoration:none;">Taskflow: A General-purpose Task-parallel Programming System</a>
</div>
</td>
</tr>
</tbody>
</table>
</div>
<!-- end header part -->
<!-- Generated by Doxygen 1.13.1 -->
<script type="text/javascript">
/* @license magnet:?xt=urn:btih:d3d9a9a6595521f9666a5e94cc830dab83b65699&dn=expat.txt MIT */
var searchBox = new SearchBox("searchBox", "search/",'.html');
/* @license-end */
</script>
<script type="text/javascript">
/* @license magnet:?xt=urn:btih:d3d9a9a6595521f9666a5e94cc830dab83b65699&dn=expat.txt MIT */
$(function() { codefold.init(0); });
/* @license-end */
</script>
<script type="text/javascript" src="menudata.js"></script>
<script type="text/javascript" src="menu.js"></script>
<script type="text/javascript">
/* @license magnet:?xt=urn:btih:d3d9a9a6595521f9666a5e94cc830dab83b65699&dn=expat.txt MIT */
$(function() {
initMenu('',true,false,'search.php','Search',true);
$(function() { init_search(); });
});
/* @license-end */
</script>
<div id="main-nav"></div>
</div><!-- top -->
<div id="side-nav" class="ui-resizable side-nav-resizable">
<div id="nav-tree">
<div id="nav-tree-contents">
<div id="nav-sync" class="sync"></div>
</div>
</div>
<div id="splitbar" style="-moz-user-select:none;"
class="ui-resizable-handle">
</div>
</div>
<script type="text/javascript">
/* @license magnet:?xt=urn:btih:d3d9a9a6595521f9666a5e94cc830dab83b65699&dn=expat.txt MIT */
$(function(){initNavTree('BenchmarkTaskflow.html',''); initResizable(true); });
/* @license-end */
</script>
<div id="doc-content">
<!-- window showing the filter options -->
<div id="MSearchSelectWindow"
onmouseover="return searchBox.OnSearchSelectShow()"
onmouseout="return searchBox.OnSearchSelectHide()"
onkeydown="return searchBox.OnSearchSelectKey(event)">
</div>
<!-- iframe showing the search results (closed by default) -->
<div id="MSearchResultsWindow">
<div id="MSearchResults">
<div class="SRPage">
<div id="SRIndex">
<div id="SRResults"></div>
<div class="SRStatus" id="Loading">Loading...</div>
<div class="SRStatus" id="Searching">Searching...</div>
<div class="SRStatus" id="NoMatches">No Matches</div>
</div>
</div>
</div>
</div>
<div><div class="header">
<div class="headertitle"><div class="title">Benchmark Taskflow</div></div>
</div><!--header-->
<div class="contents">
<div class="toc"><h3>Table of Contents</h3>
<ul>
<li class="level1">
<a href="#CompileAndRunBenchmarks">Compile and Run Benchmarks</a>
</li>
<li class="level1">
<a href="#ConfigureRunOptions">Configure Run Options</a>
<ul>
<li class="level2">
<a href="#SpecifyTheRunModel">Specify the Run Model</a>
</li>
<li class="level2">
<a href="#SpecifyTheNumberOfThreads">Specify the Number of Threads</a>
</li>
<li class="level2">
<a href="#SpecifyTheNumberOfRounds">Specify the Number of Rounds</a>
</li>
</ul>
</li>
</ul>
</div>
<div class="textblock"><h1><a class="anchor" id="CompileAndRunBenchmarks"></a>
Compile and Run Benchmarks</h1>
<p>To build the benchmark code, enable the CMake option <code>TF_BUILD_BENCHMARKS</code> to <code>ON</code> as follows:</p>
<div class="fragment"><div class="line"># under /taskflow/build</div>
<div class="line">~$ cmake ../ -DTF_BUILD_BENCHMARKS=ON</div>
<div class="line">~$ make</div>
</div><!-- fragment --><p>After you successfully build the benchmark code, you can find all benchmark instances in the <code>benchmarks/</code> folder. You can run the executable of each instance in the corresponding folder.</p>
<div class="fragment"><div class="line">~$ cd benchmarks & ls</div>
<div class="line">bench_black_scholes bench_binary_tree bench_graph_traversal ...</div>
<div class="line">~$ ./bench_graph_traversal</div>
<div class="line">|V|+|E| Runtime</div>
<div class="line"> 2 0.197</div>
<div class="line"> 842 0.198</div>
<div class="line"> 3284 0.488</div>
<div class="line"> 7288 0.774</div>
<div class="line"> ... ...</div>
<div class="line"> ... ...</div>
<div class="line"> 619802 75.135</div>
<div class="line"> 664771 77.436</div>
<div class="line"> 711200 83.957</div>
</div><!-- fragment --><p>You can display the help message by giving the option <code>--help</code>.</p>
<div class="fragment"><div class="line">~$ ./bench_graph_traversal --help</div>
<div class="line">Graph Traversal</div>
<div class="line">Usage: ./bench_graph_traversal [OPTIONS]</div>
<div class="line"> </div>
<div class="line">Options:</div>
<div class="line"> -h,--help Print this help message and exit</div>
<div class="line"> -t,--num_threads UINT number of threads (default=1)</div>
<div class="line"> -r,--num_rounds UINT number of rounds (default=1)</div>
<div class="line"> -m,--model TEXT model name tbb|omp|tf (default=tf)</div>
</div><!-- fragment --><p>We currently implement the following instances that are commonly used by the parallel computing community to evaluate the system performance.</p>
<table class="markdownTable">
<tr class="markdownTableHead">
<th class="markdownTableHeadCenter">Instance </th><th class="markdownTableHeadCenter">Description </th></tr>
<tr class="markdownTableRowOdd">
<td class="markdownTableBodyCenter">bench_binary_tree </td><td class="markdownTableBodyCenter">traverses a complete binary tree </td></tr>
<tr class="markdownTableRowEven">
<td class="markdownTableBodyCenter">bench_black_scholes </td><td class="markdownTableBodyCenter">computes option pricing with Black-Shcoles Models </td></tr>
<tr class="markdownTableRowOdd">
<td class="markdownTableBodyCenter">bench_graph_traversal </td><td class="markdownTableBodyCenter">traverses a randomly generated direct acyclic graph </td></tr>
<tr class="markdownTableRowEven">
<td class="markdownTableBodyCenter">bench_linear_chain </td><td class="markdownTableBodyCenter">traverses a linear chain of tasks </td></tr>
<tr class="markdownTableRowOdd">
<td class="markdownTableBodyCenter">bench_mandelbrot </td><td class="markdownTableBodyCenter">exploits imbalanced workloads in a Mandelbrot set </td></tr>
<tr class="markdownTableRowEven">
<td class="markdownTableBodyCenter">bench_matrix_multiplication </td><td class="markdownTableBodyCenter">multiplies two 2D matrices </td></tr>
<tr class="markdownTableRowOdd">
<td class="markdownTableBodyCenter">bench_mnist </td><td class="markdownTableBodyCenter">trains a neural network-based image classifier on the MNIST dataset </td></tr>
<tr class="markdownTableRowEven">
<td class="markdownTableBodyCenter">bench_parallel_sort </td><td class="markdownTableBodyCenter">sorts a range of items </td></tr>
<tr class="markdownTableRowOdd">
<td class="markdownTableBodyCenter">bench_reduce_sum </td><td class="markdownTableBodyCenter">sums a range of items using reduction </td></tr>
<tr class="markdownTableRowEven">
<td class="markdownTableBodyCenter">bench_wavefront </td><td class="markdownTableBodyCenter">propagates computations in a 2D grid </td></tr>
<tr class="markdownTableRowOdd">
<td class="markdownTableBodyCenter">bench_linear_pipeline </td><td class="markdownTableBodyCenter">performs pipeline parallelism on a linear chain of pipes </td></tr>
<tr class="markdownTableRowEven">
<td class="markdownTableBodyCenter">bench_graph_pipeline </td><td class="markdownTableBodyCenter">performs pipeline parallelism on a graph of pipes </td></tr>
<tr class="markdownTableRowOdd">
<td class="markdownTableBodyCenter">bench_deferred_pipeline </td><td class="markdownTableBodyCenter">performs pipeline parallelism with dependencies from future pipes </td></tr>
<tr class="markdownTableRowEven">
<td class="markdownTableBodyCenter">bench_data_pipeline </td><td class="markdownTableBodyCenter">performs pipeline parallelisms on a cache-friendly data wrapper </td></tr>
<tr class="markdownTableRowOdd">
<td class="markdownTableBodyCenter">bench_thread_pool </td><td class="markdownTableBodyCenter">uses our executor as a simple thread pool </td></tr>
<tr class="markdownTableRowEven">
<td class="markdownTableBodyCenter">bench_for_each </td><td class="markdownTableBodyCenter">performs parallel-iteration algorithms </td></tr>
<tr class="markdownTableRowOdd">
<td class="markdownTableBodyCenter">bench_scan </td><td class="markdownTableBodyCenter">performs parallel-scan algorithms </td></tr>
<tr class="markdownTableRowEven">
<td class="markdownTableBodyCenter">bench_async_task </td><td class="markdownTableBodyCenter">creates asynchronous tasks </td></tr>
<tr class="markdownTableRowOdd">
<td class="markdownTableBodyCenter">bench_fibonacci </td><td class="markdownTableBodyCenter">finds Fibonacci numbers using recursive asynchronous tasking </td></tr>
<tr class="markdownTableRowEven">
<td class="markdownTableBodyCenter">bench_nqueens </td><td class="markdownTableBodyCenter">parallelizes n-queen search using recursive asynchronous tasking </td></tr>
<tr class="markdownTableRowOdd">
<td class="markdownTableBodyCenter">bench_integrate </td><td class="markdownTableBodyCenter">parallelizes integration using recursive asynchronous tasking </td></tr>
<tr class="markdownTableRowEven">
<td class="markdownTableBodyCenter">bench_primes </td><td class="markdownTableBodyCenter">finds a range of prime numbers using parallel-reduction algorithms </td></tr>
<tr class="markdownTableRowOdd">
<td class="markdownTableBodyCenter">bench_skynet </td><td class="markdownTableBodyCenter">traverses a 10-ray tree using recursive asynchronous tasking </td></tr>
</table>
<h1><a class="anchor" id="ConfigureRunOptions"></a>
Configure Run Options</h1>
<p>We implement consistent options for each benchmark instance. Common options are:</p>
<table class="markdownTable">
<tr class="markdownTableHead">
<th class="markdownTableHeadCenter">option </th><th class="markdownTableHeadCenter">value </th><th class="markdownTableHeadCenter">function </th></tr>
<tr class="markdownTableRowOdd">
<td class="markdownTableBodyCenter"><code>-h</code> </td><td class="markdownTableBodyCenter">none </td><td class="markdownTableBodyCenter">displays the help message </td></tr>
<tr class="markdownTableRowEven">
<td class="markdownTableBodyCenter"><code>-t</code> </td><td class="markdownTableBodyCenter">integer </td><td class="markdownTableBodyCenter">configures the number of threads to run </td></tr>
<tr class="markdownTableRowOdd">
<td class="markdownTableBodyCenter"><code>-r</code> </td><td class="markdownTableBodyCenter">integer </td><td class="markdownTableBodyCenter">configures the number of rounds to run </td></tr>
<tr class="markdownTableRowEven">
<td class="markdownTableBodyCenter"><code>-m</code> </td><td class="markdownTableBodyCenter">string </td><td class="markdownTableBodyCenter">configures the baseline models to run, tbb, omp, or tf </td></tr>
</table>
<p>You can configure the benchmarking environment by giving different options.</p>
<h2><a class="anchor" id="SpecifyTheRunModel"></a>
Specify the Run Model</h2>
<p>In addition to a Taskflow-based implementation for each benchmark instance, we have implemented two baseline models using the state-of-the-art parallel programming libraries, <a href="https://www.openmp.org/">OpenMP</a> and <a href="https://github.com/oneapi-src/oneTBB">Intel TBB</a>, to measure and evaluate the performance of Taskflow. You can select different implementations by passing the option <code>-m</code>.</p>
<div class="fragment"><div class="line">~$ ./bench_graph_traversal -m tf # run the Taskflow implementation (default)</div>
<div class="line">~$ ./bench_graph_traversal -m tbb # run the TBB implementation</div>
<div class="line">~$ ./bench_graph_traversal -m omp # run the OpenMP implementation</div>
</div><!-- fragment --><h2><a class="anchor" id="SpecifyTheNumberOfThreads"></a>
Specify the Number of Threads</h2>
<p>You can configure the number of threads to run a benchmark instance by passing the option <code>-t</code>. The default value is one.</p>
<div class="fragment"><div class="line"># run the Taskflow implementation using 4 threads</div>
<div class="line">~$ ./bench_graph_traversal -m tf -t 4</div>
</div><!-- fragment --><p>Depending on your environment, you may need to use <code>taskset</code> to set the CPU affinity of the running process. This allows the OS scheduler to keep process on the same CPU(s) as long as practical for performance reason.</p>
<div class="fragment"><div class="line"># affine the process to 4 CPUs, CPU 0, CPU 1, CPU 2, and CPU 3</div>
<div class="line">~$ taskset -c 0-3 bench_graph_traversal -t 4 </div>
</div><!-- fragment --><h2><a class="anchor" id="SpecifyTheNumberOfRounds"></a>
Specify the Number of Rounds</h2>
<p>Each benchmark instance evaluates the runtime of the implementation at different problem sizes. Each problem size corresponds to one iteration. You can configure the number of rounds per iteration to average the runtime.</p>
<div class="fragment"><div class="line"># measure the %Taskflow runtime by averaging the results over 10 runs</div>
<div class="line">~$ ./bench_graph_traversal -r 10 -m tf</div>
<div class="line">|V|+|E| Runtime</div>
<div class="line"> 2 0.109 # the runtime value 0.109 is an average of 10 runs</div>
<div class="line"> 842 0.298</div>
<div class="line"> ... ...</div>
<div class="line"> 619802 73.135</div>
<div class="line"> 664771 74.436</div>
</div><!-- fragment --> </div></div><!-- contents -->
</div><!-- PageDoc -->
</div><!-- doc-content -->
<!-- HTML footer for doxygen 1.13.1-->
<!-- start footer part -->
<div id="nav-path" class="navpath"><!-- id is needed for treeview function! -->
<ul>
<li class="navelem"><a class="el" href="install.html">Building and Installing</a></li>
<li class="footer">
Maintained by <a href="https://tsung-wei-huang.github.io/">Dr. Tsung-Wei Huang</a>
—
Generated by <a href="https://www.doxygen.org/index.html"><img class="footer" src="doxygen.svg" width="104" height="31" alt="doxygen"/></a> 1.13.1
</li>
</ul>
</div>