fdurastante/amg4psblas
1
0
Fork 0
mirror of https://github.com/sfilippone/amg4psblas.git synced 2026-10-06 22:55:12 +00:00
Code Issues Packages Projects Releases Wiki Activity

mld2p4-2:

Browse Source 
Start update of documentation.
This commit is contained in:
Salvatore Filippone
2012-10-04 15:08:16 +00:00
parent bfd0d14a2a
commit ab453d36da
21 changed files with 1371 additions and 1299 deletions
Download Patch File Download Diff File Expand all files Collapse all files
docs/html/node3.html
+28 -18
View File
@@ -59,7 +59,7 @@ General Overview
<P>
The M<SMALL>ULTI-</SMALL>L<SMALL>EVEL </SMALL>D<SMALL>OMAIN </SMALL>D<SMALL>ECOMPOSITION </SMALL>P<SMALL>ARALLEL </SMALL>P<SMALL>RECONDITIONERS </SMALL>P<SMALL>ACKAGE BASED ON
</SMALL>PSBLAS (MLD2P4) provides <I>multi-level Schwarz preconditioners</I>&nbsp;[<A
HREF="node25.html#dd2_96">21</A>],
HREF="node25.html#dd2_96">22</A>],
to be used in the iterative solutions of sparse linear systems:
<BR>
<DIV ALIGN="RIGHT">
@@ -83,7 +83,8 @@ Ax=b,
where <IMG
WIDTH="18" HEIGHT="15" ALIGN="BOTTOM" BORDER="0"
SRC="img2.png"
ALT="$A$"> is a square, real or complex, sparse matrix with a symmetric sparsity pattern.
ALT="$A$"> is a square, real or complex, sparse matrix with a symmetric
sparsity pattern.
These preconditioners have the following general features:
<UL>
@@ -98,25 +99,27 @@ explicitly using any information on the geometry of the original problem (e.g. t
discretization of a PDE). The <I>smoothed aggregation</I> technique is applied
as algebraic coarsening strategy&nbsp;[<A
HREF="node25.html#BREZINA_VANEK">1</A>,<A
HREF="node25.html#VANEK_MANDEL_BREZINA">25</A>].
HREF="node25.html#VANEK_MANDEL_BREZINA">26</A>].
</LI>
</UL>
<P>
The package is written in <I>Fortran&nbsp;95</I>, following an
<I>object-oriented approach</I> through the exploitation of features
such as abstract data type creation, functional
overloading and dynamic memory management.
The parallel implementation is based
on a Single Program Multiple Data (SPMD) paradigm for distributed-memory architectures.
Single and double precision implementations of MLD2P4 are available for both the
real and the complex case, that can be used through a single interface.
Version 2.0 of the package is written in <I>Fortran&nbsp;2003</I>, following an
<I>object-oriented design</I> through the exploitation of features
such as abstract data type creation, functional overloading and
dynamic memory management.
The parallel implementation is based on a Single Program Multiple Data
(SPMD) paradigm for distributed-memory architectures. Single and
double precision implementations of MLD2P4 are available for both the
real and the complex case, that can be used through a single
interface.
<P>
MLD2P4 has been designed to implement scalable and easy-to-use multilevel preconditioners
in the context of the <I>PSBLAS (Parallel Sparse BLAS)
computational framework</I>&nbsp;[<A
HREF="node25.html#psblas_00">16</A>].
MLD2P4 has been designed to implement scalable and easy-to-use
multilevel preconditioners in the context of the <I>PSBLAS
(Parallel Sparse BLAS) computational framework</I>&nbsp;[<A
HREF="node25.html#psblas_00">17</A>,<A
HREF="node25.html#PSBLAS3">16</A>].
PSBLAS is a library originally developed to address the parallel implementation of
iterative solvers for sparse linear system, by providing basic linear algebra
operators and data management facilities for distributed sparse matrices; it
@@ -128,11 +131,11 @@ parallel sparse linear algebra kernels, to pursue goals such as performance,
portability, modularity ed extensibility in the development of the preconditioner
package. On the other hand, the implementation of MLD2P4 has led to some
revisions and extentions of the PSBLAS kernels, leading to the
recent PSBLAS 2.0 version&nbsp;[<A
PSBLAS 2.0 version&nbsp;[<A
HREF="node25.html#PSBLASGUIDE">15</A>]. The inter-process comunication required
by MLD2P4 is encapsulated into the PSBLAS routines, except few cases where
MPI&nbsp;[<A
HREF="node25.html#MPI1">22</A>] is explicitly called. Therefore, MLD2P4 can be run on any parallel
HREF="node25.html#MPI1">23</A>] is explicitly called. Therefore, MLD2P4 can be run on any parallel
machine where PSBLAS and MPI implementations are available.
<P>
@@ -147,6 +150,13 @@ On the other hand, the routines of the middle and lower layer can be used and ex
by expert users to build new versions of multi-level Schwarz preconditioners.
We provide here a description of the upper-layer routines, but not of the
medium-layer ones.
<P>
The user interface of version 2.0 is essentially identical to that of
version 1.1; the internal implementation however has been changed a
lot, and it has become much easier to extend the library by adding new
smoothers and/or solvers, thanks to the Fortran&nbsp;2003 features
exploited in the design of PSBLAS&nbsp;3.0.
<P>
This guide is organized as follows. General information on the distribution of the source code
is reported in Section&nbsp;<A HREF="node4.html#sec:distribution">2</A>, while details on the configuration
@@ -155,7 +165,7 @@ multi-level Schwarz preconditioners based on smoothed aggregation is provided
in Section&nbsp;<A HREF="node11.html#sec:background">4</A>, to help the users in choosing among the different preconditioners
implemented in MLD2P4. The basics for building and applying the preconditioners
with the Krylov solvers implemented in PSBLAS are reported in Section&nbsp;<A HREF="node14.html#sec:started">5</A>, where the
Fortran 95 codes of a few sample programs are also shown. A reference guide for
Fortran codes of a few sample programs are also shown. A reference guide for
the upper-layer routines of MLD2P4, that are the user interface, is provided
in Section&nbsp;<A HREF="node16.html#sec:userinterface">6</A>. The error handling mechanism used by the package is briefly described
in Section&nbsp;<A HREF="node23.html#sec:errors">7</A>. The copyright terms concerning the distribution and modification