Path: utzoo!utgpu!water!watmath!clyde!att!pacbell!ames!mailrus!tut.cis.ohio-state.edu!bloom-beacon!gatech!hubcap!Eugene
From: eugene@pioneer.arc.nasa.gov (Eugene N. Miya)
Newsgroups: comp.parallel
Subject: Re: parallel numerical algorithms
Message-ID: <1696@hubcap.UUCP>
Date: 24 May 88 17:10:14 GMT
Sender: fpst@hubcap.UUCP
Lines: 35
Approved: parallel@hubcap.clemson.edu

In article <1691@hubcap.UUCP> you write:
>Anyone care to try for a strict definition?

I am unable to offer a better definition, but I hope I can offer some
enlightenment based on further discussions in private others.  There is
a confusion between subtle differences in parallel/vector/Flynn's
SIMD/AND/OR parallelism what have you.  DO NOT CONFUSE PARALLEL with
SYNCHRONOUS/SYNCHRONY.  The problem which kills you are side-effects.
The early CMU people noted the major problems were: consistency,
deadlock, starvation, and exception handling.

To understand some of the problem, I offer a variation of a technique
used by the early physicists for understanding the nature of light.
They jokingly (and half-heartedly) said light behaves like particle on
MWF and like wave on TTS (Sunday, they rest).  So, DO work on parallel
algorithms or programming or machines for one week without using the
word PARALLEL.  You start to fall back on other words like CONCURRENT,
SIMULTANEOUS, MULTIPROCESS, and in time throw each of these words out.

Several different groups, I am aware of Marty Fouts and others
(independently) using words like COOPERATIVE versus COMPETITIVE
computing (why not, we have GREEDY algorithms, right? ;-)

The problem of sustaining parallel computation is sort of like
sustaining controlled thermonuclear fusion (great for short periods,
but the start up gets you).  I wish I can offer more, but I'm not smart
enough.

Another gross generalization from

--eugene miya, NASA Ames Research Center, eugene@aurora.arc.nasa.gov
  resident cynic at the Rock of Ages Home for Retired Hackers:
  "Mailers?! HA!", "If my mail does not reach you, please accept my apology."
  {uunet,hplabs,ncar,ihnp4,decwrl,allegra,tektronix}!ames!aurora!eugene
  "Send mail, avoid follow-ups.  If enough, I'll summarize."