forked from johannesgerer/jburkardt-f
-
Notifications
You must be signed in to change notification settings - Fork 1
/
task_division.html
161 lines (133 loc) · 4.44 KB
/
task_division.html
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
<html>
<head>
<title>
TASK_DIVISION - Divide T Tasks Among P Processors
</title>
</head>
<body bgcolor="#eeeeee" link="#cc0000" alink="#ff3300" vlink="#000055">
<h1 align = "center">
TASK_DIVISION <br> Divide T Tasks Among P Processors
</h1>
<hr>
<p>
<b>TASK_DIVISION</b>
is a FORTRAN90 library which
implements a simple procedure for smoothly dividing T tasks among
P processors; such a method can be useful in MPI and other parallel
environments, particularly when T is not an exact multiple of P,
and when the processors can be indexed starting from 0 or from 1.
</p>
<p>
If we have 16 tasks and 8 processors, it's easy to see that each processor
gets two tasks, and that processor 3 gets tasks 5 and 6...(unless we
index our processors starting at 0 rather than 1, but let's ignore that
possibility for the moment.)
</p>
<p>
If we have 17 tasks and 8 processors, then we can just give the last processor
the extra task, because someone has to do it. But if we have 18 tasks, we
shouldn't given everyone else 2 tasks, and the last processor 4 tasks. A
better assignment would have two of the processors working on 3 tasks, the
rest on 2.
</p>
<p>
Moreover, we'd like to do this kind of assignment in a regular enough
nonarbitrary way, so that any processor, knowning just its processor index
and the number of tasks, can work out the number of tasks it got assigned,
and even the indices of those tasks.
</p>
<p>
The TASK_DIVISION library outlines a simple way to do this, which is essentially
a "greedy" algorithm that cuts up the "task pie" a slice at a time, using rounding
to give each processor about the same number of tasks, in a way that avoids having
leftovers.
</p>
<h3 align = "center">
Licensing:
</h3>
<p>
The computer code and data files described and made available on this web page
are distributed under
<a href = "../../txt/gnu_lgpl.txt">the GNU LGPL license.</a>
</p>
<h3 align = "center">
Languages:
</h3>
<p>
<b>TASK_DIVISION</b> is available in
<a href = "../../c_src/task_division/task_division.html">a C version</a> and
<a href = "../../cpp_src/task_division/task_division.html">a C++ version</a> and
<a href = "../../f77_src/task_division/task_division.html">a FORTRAN77 version</a> and
<a href = "../../f_src/task_division/task_division.html">a FORTRAN90 version</a> and
<a href = "../../m_src/task_division/task_division.html">a MATLAB version</a>.
</p>
<h3 align = "center">
Related Data and Programs:
</h3>
<p>
<a href = "../../f_src/mpi/mpi.html">
MPI</a>,
FORTRAN90 programs which
illustrate the use of the MPI application program interface
for carrying out parallel computations in a distributed memory environment.
</p>
<h3 align = "center">
Source Code:
</h3>
<p>
<ul>
<li>
<a href = "task_division.f90">task_division.f90</a>, the source code.
</li>
<li>
<a href = "task_division.sh">task_division.sh</a>,
BASH commands to compile the source code.
</li>
</ul>
</p>
<h3 align = "center">
Examples and Tests:
</h3>
<p>
<ul>
<li>
<a href = "task_division_prb.f90">task_division_prb.f90</a>,
a sample calling program.
</li>
<li>
<a href = "task_division_prb.sh">task_division_prb.sh</a>,
BASH commands to compile and run the sample program.
</li>
<li>
<a href = "task_division_prb_output.txt">task_division_prb_output.txt</a>,
the output file.
</li>
</ul>
</p>
<h3 align = "center">
List of Routines:
</h3>
<p>
<ul>
<li>
<b>I4_DIV_ROUNDED</b> computes the rounded result of I4 division.
</li>
<li>
<b>TASK_DIVISION</b> divides tasks among processors.
</li>
<li>
<b>TIMESTAMP</b> prints the current YMDHMS date as a time stamp.
</li>
</ul>
</p>
<p>
You can go up one level to <a href = "../f_src.html">
the FORTRAN90 source codes</a>.
</p>
<hr>
<i>
Last modified on 23 October 2011.
</i>
<!-- John Burkardt -->
</body>
</html>