cannam@127
|
1 <!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN" "http://www.w3.org/TR/html4/loose.dtd">
|
cannam@127
|
2 <html>
|
cannam@127
|
3 <!-- This manual is for FFTW
|
cannam@127
|
4 (version 3.3.5, 30 July 2016).
|
cannam@127
|
5
|
cannam@127
|
6 Copyright (C) 2003 Matteo Frigo.
|
cannam@127
|
7
|
cannam@127
|
8 Copyright (C) 2003 Massachusetts Institute of Technology.
|
cannam@127
|
9
|
cannam@127
|
10 Permission is granted to make and distribute verbatim copies of this
|
cannam@127
|
11 manual provided the copyright notice and this permission notice are
|
cannam@127
|
12 preserved on all copies.
|
cannam@127
|
13
|
cannam@127
|
14 Permission is granted to copy and distribute modified versions of this
|
cannam@127
|
15 manual under the conditions for verbatim copying, provided that the
|
cannam@127
|
16 entire resulting derived work is distributed under the terms of a
|
cannam@127
|
17 permission notice identical to this one.
|
cannam@127
|
18
|
cannam@127
|
19 Permission is granted to copy and distribute translations of this manual
|
cannam@127
|
20 into another language, under the above conditions for modified versions,
|
cannam@127
|
21 except that this permission notice may be stated in a translation
|
cannam@127
|
22 approved by the Free Software Foundation. -->
|
cannam@127
|
23 <!-- Created by GNU Texinfo 5.2, http://www.gnu.org/software/texinfo/ -->
|
cannam@127
|
24 <head>
|
cannam@127
|
25 <title>FFTW 3.3.5: Distributed-memory FFTW with MPI</title>
|
cannam@127
|
26
|
cannam@127
|
27 <meta name="description" content="FFTW 3.3.5: Distributed-memory FFTW with MPI">
|
cannam@127
|
28 <meta name="keywords" content="FFTW 3.3.5: Distributed-memory FFTW with MPI">
|
cannam@127
|
29 <meta name="resource-type" content="document">
|
cannam@127
|
30 <meta name="distribution" content="global">
|
cannam@127
|
31 <meta name="Generator" content="makeinfo">
|
cannam@127
|
32 <meta http-equiv="Content-Type" content="text/html; charset=utf-8">
|
cannam@127
|
33 <link href="index.html#Top" rel="start" title="Top">
|
cannam@127
|
34 <link href="Concept-Index.html#Concept-Index" rel="index" title="Concept Index">
|
cannam@127
|
35 <link href="index.html#SEC_Contents" rel="contents" title="Table of Contents">
|
cannam@127
|
36 <link href="index.html#Top" rel="up" title="Top">
|
cannam@127
|
37 <link href="FFTW-MPI-Installation.html#FFTW-MPI-Installation" rel="next" title="FFTW MPI Installation">
|
cannam@127
|
38 <link href="Thread-safety.html#Thread-safety" rel="prev" title="Thread safety">
|
cannam@127
|
39 <style type="text/css">
|
cannam@127
|
40 <!--
|
cannam@127
|
41 a.summary-letter {text-decoration: none}
|
cannam@127
|
42 blockquote.smallquotation {font-size: smaller}
|
cannam@127
|
43 div.display {margin-left: 3.2em}
|
cannam@127
|
44 div.example {margin-left: 3.2em}
|
cannam@127
|
45 div.indentedblock {margin-left: 3.2em}
|
cannam@127
|
46 div.lisp {margin-left: 3.2em}
|
cannam@127
|
47 div.smalldisplay {margin-left: 3.2em}
|
cannam@127
|
48 div.smallexample {margin-left: 3.2em}
|
cannam@127
|
49 div.smallindentedblock {margin-left: 3.2em; font-size: smaller}
|
cannam@127
|
50 div.smalllisp {margin-left: 3.2em}
|
cannam@127
|
51 kbd {font-style:oblique}
|
cannam@127
|
52 pre.display {font-family: inherit}
|
cannam@127
|
53 pre.format {font-family: inherit}
|
cannam@127
|
54 pre.menu-comment {font-family: serif}
|
cannam@127
|
55 pre.menu-preformatted {font-family: serif}
|
cannam@127
|
56 pre.smalldisplay {font-family: inherit; font-size: smaller}
|
cannam@127
|
57 pre.smallexample {font-size: smaller}
|
cannam@127
|
58 pre.smallformat {font-family: inherit; font-size: smaller}
|
cannam@127
|
59 pre.smalllisp {font-size: smaller}
|
cannam@127
|
60 span.nocodebreak {white-space:nowrap}
|
cannam@127
|
61 span.nolinebreak {white-space:nowrap}
|
cannam@127
|
62 span.roman {font-family:serif; font-weight:normal}
|
cannam@127
|
63 span.sansserif {font-family:sans-serif; font-weight:normal}
|
cannam@127
|
64 ul.no-bullet {list-style: none}
|
cannam@127
|
65 -->
|
cannam@127
|
66 </style>
|
cannam@127
|
67
|
cannam@127
|
68
|
cannam@127
|
69 </head>
|
cannam@127
|
70
|
cannam@127
|
71 <body lang="en" bgcolor="#FFFFFF" text="#000000" link="#0000FF" vlink="#800080" alink="#FF0000">
|
cannam@127
|
72 <a name="Distributed_002dmemory-FFTW-with-MPI"></a>
|
cannam@127
|
73 <div class="header">
|
cannam@127
|
74 <p>
|
cannam@127
|
75 Next: <a href="Calling-FFTW-from-Modern-Fortran.html#Calling-FFTW-from-Modern-Fortran" accesskey="n" rel="next">Calling FFTW from Modern Fortran</a>, Previous: <a href="Multi_002dthreaded-FFTW.html#Multi_002dthreaded-FFTW" accesskey="p" rel="prev">Multi-threaded FFTW</a>, Up: <a href="index.html#Top" accesskey="u" rel="up">Top</a> [<a href="index.html#SEC_Contents" title="Table of contents" rel="contents">Contents</a>][<a href="Concept-Index.html#Concept-Index" title="Index" rel="index">Index</a>]</p>
|
cannam@127
|
76 </div>
|
cannam@127
|
77 <hr>
|
cannam@127
|
78 <a name="Distributed_002dmemory-FFTW-with-MPI-1"></a>
|
cannam@127
|
79 <h2 class="chapter">6 Distributed-memory FFTW with MPI</h2>
|
cannam@127
|
80 <a name="index-MPI"></a>
|
cannam@127
|
81
|
cannam@127
|
82 <a name="index-parallel-transform-1"></a>
|
cannam@127
|
83 <p>In this chapter we document the parallel FFTW routines for parallel
|
cannam@127
|
84 systems supporting the MPI message-passing interface. Unlike the
|
cannam@127
|
85 shared-memory threads described in the previous chapter, MPI allows
|
cannam@127
|
86 you to use <em>distributed-memory</em> parallelism, where each CPU has
|
cannam@127
|
87 its own separate memory, and which can scale up to clusters of many
|
cannam@127
|
88 thousands of processors. This capability comes at a price, however:
|
cannam@127
|
89 each process only stores a <em>portion</em> of the data to be
|
cannam@127
|
90 transformed, which means that the data structures and
|
cannam@127
|
91 programming-interface are quite different from the serial or threads
|
cannam@127
|
92 versions of FFTW.
|
cannam@127
|
93 <a name="index-data-distribution"></a>
|
cannam@127
|
94 </p>
|
cannam@127
|
95
|
cannam@127
|
96 <p>Distributed-memory parallelism is especially useful when you are
|
cannam@127
|
97 transforming arrays so large that they do not fit into the memory of a
|
cannam@127
|
98 single processor. The storage per-process required by FFTW’s MPI
|
cannam@127
|
99 routines is proportional to the total array size divided by the number
|
cannam@127
|
100 of processes. Conversely, distributed-memory parallelism can easily
|
cannam@127
|
101 pose an unacceptably high communications overhead for small problems;
|
cannam@127
|
102 the threshold problem size for which parallelism becomes advantageous
|
cannam@127
|
103 will depend on the precise problem you are interested in, your
|
cannam@127
|
104 hardware, and your MPI implementation.
|
cannam@127
|
105 </p>
|
cannam@127
|
106 <p>A note on terminology: in MPI, you divide the data among a set of
|
cannam@127
|
107 “processes” which each run in their own memory address space.
|
cannam@127
|
108 Generally, each process runs on a different physical processor, but
|
cannam@127
|
109 this is not required. A set of processes in MPI is described by an
|
cannam@127
|
110 opaque data structure called a “communicator,” the most common of
|
cannam@127
|
111 which is the predefined communicator <code>MPI_COMM_WORLD</code> which
|
cannam@127
|
112 refers to <em>all</em> processes. For more information on these and
|
cannam@127
|
113 other concepts common to all MPI programs, we refer the reader to the
|
cannam@127
|
114 documentation at <a href="http://www.mcs.anl.gov/research/projects/mpi/">the MPI home
|
cannam@127
|
115 page</a>.
|
cannam@127
|
116 <a name="index-MPI-communicator"></a>
|
cannam@127
|
117 <a name="index-MPI_005fCOMM_005fWORLD"></a>
|
cannam@127
|
118 </p>
|
cannam@127
|
119
|
cannam@127
|
120 <p>We assume in this chapter that the reader is familiar with the usage
|
cannam@127
|
121 of the serial (uniprocessor) FFTW, and focus only on the concepts new
|
cannam@127
|
122 to the MPI interface.
|
cannam@127
|
123 </p>
|
cannam@127
|
124 <table class="menu" border="0" cellspacing="0">
|
cannam@127
|
125 <tr><td align="left" valign="top">• <a href="FFTW-MPI-Installation.html#FFTW-MPI-Installation" accesskey="1">FFTW MPI Installation</a>:</td><td> </td><td align="left" valign="top">
|
cannam@127
|
126 </td></tr>
|
cannam@127
|
127 <tr><td align="left" valign="top">• <a href="Linking-and-Initializing-MPI-FFTW.html#Linking-and-Initializing-MPI-FFTW" accesskey="2">Linking and Initializing MPI FFTW</a>:</td><td> </td><td align="left" valign="top">
|
cannam@127
|
128 </td></tr>
|
cannam@127
|
129 <tr><td align="left" valign="top">• <a href="2d-MPI-example.html#g_t2d-MPI-example" accesskey="3">2d MPI example</a>:</td><td> </td><td align="left" valign="top">
|
cannam@127
|
130 </td></tr>
|
cannam@127
|
131 <tr><td align="left" valign="top">• <a href="MPI-Data-Distribution.html#MPI-Data-Distribution" accesskey="4">MPI Data Distribution</a>:</td><td> </td><td align="left" valign="top">
|
cannam@127
|
132 </td></tr>
|
cannam@127
|
133 <tr><td align="left" valign="top">• <a href="Multi_002ddimensional-MPI-DFTs-of-Real-Data.html#Multi_002ddimensional-MPI-DFTs-of-Real-Data" accesskey="5">Multi-dimensional MPI DFTs of Real Data</a>:</td><td> </td><td align="left" valign="top">
|
cannam@127
|
134 </td></tr>
|
cannam@127
|
135 <tr><td align="left" valign="top">• <a href="Other-Multi_002ddimensional-Real_002ddata-MPI-Transforms.html#Other-Multi_002ddimensional-Real_002ddata-MPI-Transforms" accesskey="6">Other Multi-dimensional Real-data MPI Transforms</a>:</td><td> </td><td align="left" valign="top">
|
cannam@127
|
136 </td></tr>
|
cannam@127
|
137 <tr><td align="left" valign="top">• <a href="FFTW-MPI-Transposes.html#FFTW-MPI-Transposes" accesskey="7">FFTW MPI Transposes</a>:</td><td> </td><td align="left" valign="top">
|
cannam@127
|
138 </td></tr>
|
cannam@127
|
139 <tr><td align="left" valign="top">• <a href="FFTW-MPI-Wisdom.html#FFTW-MPI-Wisdom" accesskey="8">FFTW MPI Wisdom</a>:</td><td> </td><td align="left" valign="top">
|
cannam@127
|
140 </td></tr>
|
cannam@127
|
141 <tr><td align="left" valign="top">• <a href="Avoiding-MPI-Deadlocks.html#Avoiding-MPI-Deadlocks" accesskey="9">Avoiding MPI Deadlocks</a>:</td><td> </td><td align="left" valign="top">
|
cannam@127
|
142 </td></tr>
|
cannam@127
|
143 <tr><td align="left" valign="top">• <a href="FFTW-MPI-Performance-Tips.html#FFTW-MPI-Performance-Tips">FFTW MPI Performance Tips</a>:</td><td> </td><td align="left" valign="top">
|
cannam@127
|
144 </td></tr>
|
cannam@127
|
145 <tr><td align="left" valign="top">• <a href="Combining-MPI-and-Threads.html#Combining-MPI-and-Threads">Combining MPI and Threads</a>:</td><td> </td><td align="left" valign="top">
|
cannam@127
|
146 </td></tr>
|
cannam@127
|
147 <tr><td align="left" valign="top">• <a href="FFTW-MPI-Reference.html#FFTW-MPI-Reference">FFTW MPI Reference</a>:</td><td> </td><td align="left" valign="top">
|
cannam@127
|
148 </td></tr>
|
cannam@127
|
149 <tr><td align="left" valign="top">• <a href="FFTW-MPI-Fortran-Interface.html#FFTW-MPI-Fortran-Interface">FFTW MPI Fortran Interface</a>:</td><td> </td><td align="left" valign="top">
|
cannam@127
|
150 </td></tr>
|
cannam@127
|
151 </table>
|
cannam@127
|
152
|
cannam@127
|
153 <hr>
|
cannam@127
|
154 <div class="header">
|
cannam@127
|
155 <p>
|
cannam@127
|
156 Next: <a href="Calling-FFTW-from-Modern-Fortran.html#Calling-FFTW-from-Modern-Fortran" accesskey="n" rel="next">Calling FFTW from Modern Fortran</a>, Previous: <a href="Multi_002dthreaded-FFTW.html#Multi_002dthreaded-FFTW" accesskey="p" rel="prev">Multi-threaded FFTW</a>, Up: <a href="index.html#Top" accesskey="u" rel="up">Top</a> [<a href="index.html#SEC_Contents" title="Table of contents" rel="contents">Contents</a>][<a href="Concept-Index.html#Concept-Index" title="Index" rel="index">Index</a>]</p>
|
cannam@127
|
157 </div>
|
cannam@127
|
158
|
cannam@127
|
159
|
cannam@127
|
160
|
cannam@127
|
161 </body>
|
cannam@127
|
162 </html>
|