Newsgroups: comp.parallel
Path: tk
From: Tommy Kelly <tk@dcs.ed.ac.uk>
Subject: Re: Seeking suggestions on simulation tools
Sender: UseNet News Admin <cnews@dcs.ed.ac.uk>
Reply-To: Tommy Kelly <tk@dcs.ed.ac.uk>
Organization: Laboratory for the Foundations of Computer Science, Edinburgh U
References: <1993May27.175802.29144@hubcap.clemson.edu> <1993May31.143339.17377@hubcap.clemson.edu>
Date: Mon, 31 May 1993 21:31:35 GMT
Lines: 45
Apparently-To: comp-parallel@uknet.ac.uk

Is this an FAQ?  I've seen variations of it pop up several times now.

loren@cis.ohio-state.edu (Loren Schwiebert) writes:

>I am interested in doing a simulation of message routing on a variety
>of topologies and using a variety of routing algorithms.  I am
>considering using CSIM, but was wondering if anyone had experience
>with/suggestions for other simulation languages or tools.

You could check out:

CACI SIMSCRIPT II.5
	This is available on DOS and on Suns (at least).  A colleague
	of mine has used this extensively for just the sort of thing
	you mention.

SIM++ (from JSI in Canada)
	I've used this quite a bit.  It is a set of C++ libraries and various
	bells and whistles.  It was the C++ aspect that appealed to me.
	But it also has the advantage that it can
	be run in parallel on a network of Suns or on a transputer
	machine (if you have one).

  Both of the above are commercial products.

Other possibilities are:

AWESIME (available from Boulder Uni, Colorado?)
	I think this is C++ based too.  Its also free.
	Also, I know of a bolt-on called GENESIS which I think 
	was used to make lower-level hardware simulation easier.
	
DEMOS (Discrete Event Modelling On Simula)
	As the name suggests, this is based on Simula 67.  Old as the
	latter may be, it has apparantly weathered well, and is arguable
	easier to use than C++.
	Here at Edinburgh some folks have written another layer, to sit
	on top of DEMOS (which sits on top of SIMULA) to make parallel
	computer simulations easier (this layer is called MIMD).

I hesitate to suggest any one over the other.  In answering the very
same question to a student recently I replied that unless the application
was particularly heavy, or peculiar, then any one of the above would do.

tommy