Path: utzoo!attcan!uunet!husc6!uwvax!oddjob!ncar!gatech!hubcap!fpst
From: fpst@hubcap.clemson.edu (Steve Stevenson-Moderator)
Newsgroups: comp.parallel
Subject: Re: parallel numerical algorithms
Message-ID: <1691@hubcap.UUCP>
Date: 24 May 88 12:52:19 GMT
Sender: fpst@hubcap.UUCP
Lines: 16
Approved: parallel@hubcap.clemson.edu

From article <1684@hubcap.UUCP>, by Eugene N. Miya <eugene@pioneer.arc.nasa.gov>:
> 
> If we are going to make a distinction between vector and parallel, it had
> better have better basis for making that distinction.
> 

I agree.  It seems to me that a tentative, first cut distinction
is SIMD and MIMD but one has to be very careful.  Example: a possible
counter is that NCUBE does a parallel algorithm (in MIMD sense) 'cause
it does not have vector coprocessors.  The FPS T, on the other hand,
could do both.

It seems to me that the essentials have to do with some reformulation
of linear algebra and some agreement that there is an equivalence.

Anyone care to try for a strict definition?