home .. forth .. misc mail list archive ..

Re: Hardware/Software


Wayne,

What do I think?  I think you've completely missed the boat:

The sole reason that MISC processors are of interest is that they  
minimize the real-estate required to implement a processor.

Why is this of interest?  Because you can put a maximum number of  
processors in minimal real-estate [read:  on a single chip].  This  
enables fine-grained computing.

If you need to worry about interfacing I/O devices to individual  
MISC cells, you have the wrong architecture.

Think about it.
				Sherwin Gooch


Begin forwarded message:

Resent-Date: Sun, 28 Jan 1996 22:59:20 -0500
From: Wayne Morellini <waynem1@cq-pan.cqu.edu.au>
Subject: Hardware/Software
To: misc@pisa.rockefeller.edu (misc)
Date: Mon, 29 Jan 1996 14:00:24 +1000 (EST)
Cc: waynem1@cq-pan.cqu.edu.au (Wayne Morellini)
X-Mailer: ELM [version 2.4 PL23]
Resent-From: MISC@pisa.rockefeller.edu
X-Mailing-List: <MISC@pisa.rockefeller.edu> archive/latest/319
X-Loop: MISC@pisa.rockefeller.edu
Resent-Sender: MISC-request@pisa.rockefeller.edu

Hi Miscers

I am interested in some discussion of development of products for  
the Misc
designs.

I am writing this because of an obvious lack of support components  
for the
Misc processors, and the possibility that people maybe turned off  
of using
or doing something with the chip because they would have to start from
scratch and also the possibility of duplication of effort.

I propose that we have some discussion on what is needed to develop
products using the processors, and as people wish they may go ahead and
independently develop these components and offer them for resale/license
to other individuals for inclusion in there designs.

I see the need for such a library of components so that:

a)  People don't keep re-inventing the component for each design (though
there maybe individual competing designs of each component, for  
choice, but
at least we won't have to do it every time).

b)  Extensively cut down development time/cost of finished products  
by using
components.

c) Third party market support for the processors, which is vital for the
confidence of many developers who may wish to adopt the design.

d)  It helps promote Misc processors (our favorite).

e)  Gives you an interesting hobby in doing designs/business and

f)  Gives you some in site into this particular type of business  
and design
work, leading to the possibility of making a go of it and maybe  
some money.

g)  Personal satisfaction, and if I keep on going on I'll start sounding
new age:).


Look at this way: a lot of products might need I/O, disk drives, Graphic
chips, PCMCIA V2.0 compliance, hardware debuggers, software debuggers,
languages, data analysis, aquistion, databases etc, whatever you  
can think
of, for my area certain things are prefferable, and in other areas,  
others.


I propose that we discuss a market area, say embedded controllers,  
and see
what we need in it, whats prefered and what people want.  For embedded
controllers (only as an example) people might include that the MicroChip
model is the best and that we should have some products like the  
Basic Stamp
etc, and break it down.  Once we finish one area we go onto another  
area.
If we finish all areas we can take a look at the first area again  
to see how
it is doing, and so on.  All this time people who are interested  
may start
developing a component and tell the group that they are, so that  
not everybody
goes to do it, and we move on.  I am talking about both hardware and
software components (with software accompanying the hardware).   
individuals
could even approach places like Mr Tings Offette Enterprises, or the
processor owners to distribute their components.

The only really solid things that I would suggest is that we  
standardise on
two types of designs, one a custom compact fast version that maybe single
tasking, single user (etc).  The second a design that would be  
suitable for
multi-taking environments, with re-entrant software, totally  
relocatable in
software and hardware, in an object oriented fashion.  This is so the
component (hardware and software) can be located in any part of  
memory/I/O
it has been assigned and be encapsulated with the host system  
software (so
the programs merely see an software interface), runs with all other  
software
without effecting there execution path, and allows more them one  
session in
itself.

As far as myself goes, I feel I can reveal that I am working on an mini
operating system in the fashion Taos and Java, and I yet might decide to
port one of these competing products instead, so don't bank on it yet.

I suggest that we start off with a list of market areas suitable for the
Misc processors (Mup21, F21, P32? and what ever else).

Micro-controller
Networking
AI
PDA's/Organizers
Video Games/Portable
Computers
Internet
DSP
etc

What do you think?
Do you have any suggestions to add to the list?



Wayne Morellini
Microsun

waynem1@cq-pan.cqu.edu.au

--