Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!uwm.edu!ux1.cso.uiuc.edu!uicsrd.csrd.uiuc.edu!hsu
From: hsu@uicsrd.csrd.uiuc.edu (William Tsun-Yuk Hsu)
Newsgroups: comp.arch
Subject: Re: parallel systems
Message-ID: <1989Oct24.210602.2616@ux1.cso.uiuc.edu>
Date: 24 Oct 89 21:06:02 GMT
References: <35825@lll-winken.LLNL.GOV> <20336@princeton.Princeton.EDU> <7651@bunny.GTE.COM> <1989Oct23.152120.25967@cs.rochester.edu> <6655@ficc.uu.net>
Sender: paul@ux1.cso.uiuc.edu (Paul Pomes)
Reply-To: hsu@uicsrd.csrd.uiuc.edu (William Tsun-Yuk Hsu)
Organization: Center for Supercomputing Research and Development, Univ. of Illinois
Lines: 15

In article <6655@ficc.uu.net> peter@ficc.uu.net (Peter da Silva) writes:
>
>Also, the job of programming a shared memory system is a lot harder than
>programming a system with messages as the communication medium. 

Umm, have you tried? 

When I took a parallel programming class a couple years ago, everybody
griped about the problems of coding Gaussian elimination for the
hypercube. Hardly anybody griped about implementing it for the
Sequent balance. (This single example of course doesn't prove
anything, but I would like to see some examples of problems that
are clearly easier to code for a message-passing system.)

Bill