CMSSL routine for solving 2D finite difference?

Thread View: comp.sys.super

2 messages

[T] Back to Threads [V] Tree View [G] Group Articles

2 total messages Started by Jay Martinez Sun, 06 Sep 1992 05:06

#143

Author: Jay Martinez
Date: Sun, 06 Sep 1992 05:06

34 lines
2035 bytes


   I am trying to do a two-dimensional finite difference formulation
   on a CM-200.  For the resulting system of equations Ax=B, I get an
   A maxtrix that looks like this:

    
               |  c    d              e                 |
               |                                        |
               |  b    c    d              e            |
               |                                        |
               |       b    c    d              e       |
               |                                        |
               |           b    c    d               e  |
               |                                        |
               |  a              b    c    d            |
               |                                        |
               |       a              b    c    d       |
               |                                        |
               |            a              b    c    d  |
               |                                        |
               |                 a             b    c   |

   where the space between the centre tridiagonal band and the outer
   bands is two.  Does anyone know of a CMSSL routine that will solve
   such a system?  Or is there anyway to transform this system into
   one that can be solved by another CMSSL sparse system routine?
   The CMSSL routine gen_penta_diag doesn't seem to have a provision
   for spaces between bands.  I would appreciate any suggestions.

   Jay Jansen Martinez (jamartin@NMSU.Edu)
   
   Department of Electrical and Computer Engineering
   New Mexico State University

Re: CMSSL routine for solving 2D finite difference?

#145

Author: Robert Krawitz
Date: Thu, 10 Sep 1992 15:30

26 lines
1620 bytes

In article <JAMARTIN.92Sep5220655@gauss.nmsu.edu> jamartin@nmsu.edu (Jay Martinez) writes:

      I am trying to do a two-dimensional finite difference formulation
      on a CM-200.  For the resulting system of equations Ax=B, I get an
      A maxtrix that looks like this:

Lennart Johnsson (Director of Computational Sciences) made the following
comments:

To: Robert L Krawitz <rlk@Think.COM>
Subject: Re: CMSSL routine for solving 2D finite difference? 
Date: Tue, 08 Sep 92 22:35:22 EDT
From: Lennart Johnsson <johnsson@Think.COM>

	They can solve it iteratively using the grid-sparse matrix-vector
	routine, or embed it in a block tridiagonal system (not necessarily 
	very efficient)

	The performance of an iterative solver should be further enhanced
	by the Stencil compiler in 3.1.
--
ames >>>>>>>>>  |	Robert Krawitz <rlk@think.com>	245 First St.
bloom-beacon >  |think!rlk				Cambridge, MA  02142
harvard >>>>>>  .	Thinking Machines Corp.		(617)234-2116

Member of the League for Programming Freedom -- write league@prep.ai.mit.edu

Thread Navigation

This is a paginated view of messages in the thread with full content displayed inline.

Messages are displayed in chronological order, with the original post highlighted in green.

Use pagination controls to navigate through all messages in large threads.

Back to All Threads