# Parallel Algorithms for Matrix Multiplication

Can anyone tell me where I can find a list of different parallel algorithms (or source codes) for matrix multiplication ?
###### Who is Participating?
I wear a lot of hats...

"The solutions and answers provided on Experts Exchange have been extremely helpful to me over the last few years. I wear a lot of hats - Developer, Database Administrator, Help Desk, etc., so I know a lot of things but not a lot about one thing. Experts Exchange gives me answers from people who do know a lot about one thing, in a easy to use platform." -Todd S.

Author Commented:
Edited text of question
0
Commented:
0
Author Commented:
Thanks ozo. But is there a site which explains several algorithms?
I'm actually looking for recursive ones, but it seems like there are not many in the web.
0
Commented:
Is there a particular parallel architecture you want to run on?
0
Author Commented:
I would be coding the algo in mpi, and would like to look for a few parallel recursive methods to evaluate.
I've thought of a simple tree-structured one, which divides the matrix into 3 sets of rows, 2 for the children and one to self-compute. But then again, it needs further improvement.
0
Commented:
Take a look into http://www.cs.sandia.gov/~bahendr/lin_alg.html

There you will find asome papers dealing on linear algebra and
parallel algorithms....

Best regards
.... Taliesin
0
Author Commented:
Thanks Taliesin, but the papers do not exactly fit my requirements, nonetheless, they help.
0
Commented:

1) A Three-dimensional Approach to Parallel Matrix Multiplication by R. C. Agarwal, S. M. Balle, F. G. Gustavson, M. Joshi, and P. Palkar.
Implementation of a 3D algorithm on an IBM SP2. :

(with nice references!)

2) A Scalable Parallel Strassen's Matrix Multiply Algorithm
for Distributed Memory Computers by Qingshan Luo and John B. Drake.
Parallel algorithm and implementation on 128-processor Intel iPSC. :
http://www.epm.ornl.gov/~bbd/pubs/stras5.ps

3) Numerical algorithms for supercomputers:
http://www.math.ruu.nl/people/bisselin/nas.html

My choice is N°1, Strassen method, modified by Winograd, and
working as 3D matrices, not 2D....

pretty cool ... ;)

0

Experts Exchange Solution brought to you by