| 1 |
<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2 Final//EN"> |
| 2 |
|
| 3 |
<!--Converted with LaTeX2HTML 2019.2 (Released June 5, 2019) --> |
| 4 |
<HTML lang="EN"> |
| 5 |
<HEAD> |
| 6 |
<TITLE>3.1 Understanding Parallelism</TITLE> |
| 7 |
<META NAME="description" CONTENT="3.1 Understanding Parallelism"> |
| 8 |
<META NAME="keywords" CONTENT="user_guide"> |
| 9 |
<META NAME="resource-type" CONTENT="document"> |
| 10 |
<META NAME="distribution" CONTENT="global"> |
| 11 |
|
| 12 |
<META HTTP-EQUIV="Content-Type" CONTENT="text/html; charset=utf-8"> |
| 13 |
<META NAME="viewport" CONTENT="width=device-width, initial-scale=1.0"> |
| 14 |
<META NAME="Generator" CONTENT="LaTeX2HTML v2019.2"> |
| 15 |
|
| 16 |
<LINK REL="STYLESHEET" HREF="user_guide.css"> |
| 17 |
|
| 18 |
<LINK REL="next" HREF="node19.html"> |
| 19 |
<LINK REL="previous" HREF="node17.html"> |
| 20 |
<LINK REL="next" HREF="node19.html"> |
| 21 |
</HEAD> |
| 22 |
|
| 23 |
<BODY > |
| 24 |
<!--Navigation Panel--> |
| 25 |
<A |
| 26 |
HREF="node19.html"> |
| 27 |
<IMG WIDTH="37" HEIGHT="24" ALT="next" SRC="next.png"></A> |
| 28 |
<A |
| 29 |
HREF="node17.html"> |
| 30 |
<IMG WIDTH="26" HEIGHT="24" ALT="up" SRC="up.png"></A> |
| 31 |
<A |
| 32 |
HREF="node17.html"> |
| 33 |
<IMG WIDTH="63" HEIGHT="24" ALT="previous" SRC="prev.png"></A> |
| 34 |
<A ID="tex2html221" |
| 35 |
HREF="node1.html"> |
| 36 |
<IMG WIDTH="65" HEIGHT="24" ALT="contents" SRC="contents.png"></A> |
| 37 |
<BR> |
| 38 |
<B> Next:</B> <A |
| 39 |
HREF="node19.html">3.2 Running on parallel</A> |
| 40 |
<B> Up:</B> <A |
| 41 |
HREF="node17.html">3 Parallelism</A> |
| 42 |
<B> Previous:</B> <A |
| 43 |
HREF="node17.html">3 Parallelism</A> |
| 44 |
<B> <A ID="tex2html222" |
| 45 |
HREF="node1.html">Contents</A></B> |
| 46 |
<BR> |
| 47 |
<BR> |
| 48 |
<!--End of Navigation Panel--> |
| 49 |
|
| 50 |
<H2><A ID="SECTION00041000000000000000"> |
| 51 |
3.1 Understanding Parallelism</A> |
| 52 |
</H2> |
| 53 |
|
| 54 |
<P> |
| 55 |
Two different parallelization paradigms are currently implemented |
| 56 |
in Q<SMALL>UANTUM </SMALL>ESPRESSO: |
| 57 |
|
| 58 |
<OL> |
| 59 |
<LI><EM>Message-Passing (MPI)</EM>. A copy of the executable runs |
| 60 |
on each CPU; each copy lives in a different world, with its own |
| 61 |
private set of data, and communicates with other executables only |
| 62 |
via calls to MPI libraries. MPI parallelization requires compilation |
| 63 |
for parallel execution, linking with MPI libraries, execution using |
| 64 |
a launcher program (depending upon the specific machine). The number of CPUs used |
| 65 |
is specified at run-time either as an option to the launcher or |
| 66 |
by the batch queue system. |
| 67 |
</LI> |
| 68 |
<LI><EM>OpenMP</EM>. A single executable spawn subprocesses |
| 69 |
(threads) that perform in parallel specific tasks. |
| 70 |
OpenMP can be implemented via compiler directives (<EM>explicit</EM> |
| 71 |
OpenMP) or via <EM>multithreading</EM> libraries (<EM>library</EM> OpenMP). |
| 72 |
Explicit OpenMP require compilation for OpenMP execution; |
| 73 |
library OpenMP requires only linking to a multithreading |
| 74 |
version of mathematical libraries, e.g.: |
| 75 |
ESSLSMP, ACML_MP, MKL (the latter is natively multithreading). |
| 76 |
The number of threads is specified at run-time in the environment |
| 77 |
variable OMP_NUM_THREADS. |
| 78 |
</LI> |
| 79 |
</OL> |
| 80 |
|
| 81 |
<P> |
| 82 |
MPI is the well-established, general-purpose parallelization. |
| 83 |
In Q<SMALL>UANTUM </SMALL>ESPRESSO several parallelization levels, specified at run-time |
| 84 |
via command-line options to the executable, are implemented |
| 85 |
with MPI. This is your first choice for execution on a parallel |
| 86 |
machine. |
| 87 |
|
| 88 |
<P> |
| 89 |
The support for explicit OpenMP is steadily improving. |
| 90 |
Explicit OpenMP can be used together with MPI and also |
| 91 |
together with library OpenMP. Beware |
| 92 |
conflicts between the various kinds of parallelization! |
| 93 |
If you don't know how to run MPI processes |
| 94 |
and OpenMP threads in a controlled manner, forget about mixed |
| 95 |
OpenMP-MPI parallelization. |
| 96 |
|
| 97 |
<P> |
| 98 |
<HR> |
| 99 |
<!--Navigation Panel--> |
| 100 |
<A |
| 101 |
HREF="node19.html"> |
| 102 |
<IMG WIDTH="37" HEIGHT="24" ALT="next" SRC="next.png"></A> |
| 103 |
<A |
| 104 |
HREF="node17.html"> |
| 105 |
<IMG WIDTH="26" HEIGHT="24" ALT="up" SRC="up.png"></A> |
| 106 |
<A |
| 107 |
HREF="node17.html"> |
| 108 |
<IMG WIDTH="63" HEIGHT="24" ALT="previous" SRC="prev.png"></A> |
| 109 |
<A ID="tex2html221" |
| 110 |
HREF="node1.html"> |
| 111 |
<IMG WIDTH="65" HEIGHT="24" ALT="contents" SRC="contents.png"></A> |
| 112 |
<BR> |
| 113 |
<B> Next:</B> <A |
| 114 |
HREF="node19.html">3.2 Running on parallel</A> |
| 115 |
<B> Up:</B> <A |
| 116 |
HREF="node17.html">3 Parallelism</A> |
| 117 |
<B> Previous:</B> <A |
| 118 |
HREF="node17.html">3 Parallelism</A> |
| 119 |
<B> <A ID="tex2html222" |
| 120 |
HREF="node1.html">Contents</A></B> |
| 121 |
<!--End of Navigation Panel--> |
| 122 |
|
| 123 |
</BODY> |
| 124 |
</HTML> |