Parallel Matrix Multiplication

Article View: comp.sys.transputer

Article #1216

Back to comp.sys.transputer Previous Next

#1216

From: Ye Huang
Date: Mon, 12 Oct 1992 00:53

29 lines
1516 bytes

Hello There,

We have a Meiko transputer system. I am trying different mappings to
perform the MATRICES MULTIPLICATION on it. 

Our configuration has 32 processors. I have the coded for mesh, torus.
But the speedup is not quite close to the theoretical results (and
even some paper reported results).

My guessing is that probably I did something inaproporate in the
implementation of them. 

I would appreciate to receive some your implementation of this
problem as pointers to benchmark my implementation. Preferabl on Meiko
since we only have meiko system here.

I will post a summary along with my final version of the problem.

Cheers.

---Thomas
--
: Ye Huang                                      
: e-mail: yhuang@bishop.ics.hawaii.edu                  
: phone:  808-956-3490(W) 808-946-0969(R)            
: s-mail: software engineering research lab., 
:         ICS Dept., UH, Keller301, HNL, HI 96822   
:........................................................................

Technical Details

Message-ID: <YHUANG.92Oct11145315@bishop.ics.hawaii.edu>

Path:

arc1-ca.pugleaf.net!sparky!uunet!convex!darwin.sura.net!spool.mu.edu!sdd.hp.com!elroy.jpl.nasa.gov!ames!news.hawaii.edu!news!yhuang

Show Full Headers

Complete Article Headers

Newsgroups: comp.sys.transputer
Path: arc1-ca.pugleaf.net!sparky!uunet!convex!darwin.sura.net!spool.mu.edu!sdd.hp.com!elroy.jpl.nasa.gov!ames!news.hawaii.edu!news!yhuang
From: yhuang@bishop.ics.hawaii.edu (Ye Huang )
Subject: Parallel Matrix Multiplication 
Message-ID: <YHUANG.92Oct11145315@bishop.ics.hawaii.edu>
Sender: root@news.Hawaii.Edu (News Service)
Nntp-Posting-Host: bishop.ics.hawaii.edu
Organization: SRA, Inc.
Distribution: comp.sys.transputer
Date: Mon, 12 Oct 1992 00:53:15 GMT
Lines: 29