Newsgroups: comp.parallel.pvm,comp.parallel
From: rcarter@best.com (Russell Carter)
Subject: Re: [Q]:Parallel Matrix Multiplication
Organization: Best Internet Communications
Date: Thu, 9 Nov 1995 19:59:46 GMT
Message-ID: <47tmji$q7g@shellx.best.com>

In article <47n2tb$lpm@post.tau.ac.il>,
Pick Daniel <pick@lune.math.tau.ac.il> wrote:
>In article <475gre$spf@diable.upc.es> Antonio Scotti <scotti@centauro.upc.es> writes:
>>Hi,
>>does anybody know how many and which algorithms for parallel matrix
>>multiplication are there around?
>>Does anybody have references to some relevant literature?
>>
>>Thanks
>>-- Antonio
>>
>>
>     I think at the moment you write your own :)  Most of the papers that I'm
>aware of on this subject are for either computing the eigenvalues of the
>matrix, or for sparse matrices.  The compilers for the Intel Paragon have
>built-in global routines that allow you to compute the result of a computation
>across many processors, and store the result in one.  This makes it easy 
>to write your own code to perform fast matrix multiplication.
>
>                                                 --Dan 8^)
>

Always interesting to do, but you might have a look at the SUMMA
code and papers.  Search http://www.netlib.org.  I built it up yesterday,
runs fine with MPICH on a Pentium workstation cluster.

Russell Carter
http://www.geli.com


>------------------------------------------------------------------------
>Daniel Pick
>
>e-mail: pickd@cgs.edu                  
>snail: Visiting Fulbright scholar
>       Dept. of Applied Mathematics       (This space under construction.)
>       Tel Aviv University
>       Ramat Aviv 69978 Israel
>

