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

Re: never enough


Dear MISC readers:

Nathaniel Downes wrote:

> > We are currently (in one project) looking at reducing the design by
> > making I/O register based (no mems) and integrating a small amount
> > of memory on chip and multiple nodes connected together on a chip.
> > With wafer scale integration it looks like the upper limit is
> > somewhere around 15 million Forth mips from a "chip" in production
> > silicon today.  Of course I personally don't have the money to
> > do the design so "feasability" depends entirely on finding
> > the money like most other things.
> 
> I got more than that.

Yes, me too.  Richard Westmorrland sent me a reference to
the nanofab at Penn State.  htpp://www.nanofab.psu.edu

I was very impressed.  They are offing the use of their
nanofab for the R&D community for micro and nano electronics,
opticial, and micro-electromechanical (MEMS) designs with
a process in 20 nanometers, .02 micron.  We have been looking
at going from .6u to .35 or maybe .25 if we are lucky.  I
didn't know anything like .02 micron might be available.

Of course designs would have to change since line delay
becomes more signifigant than gate delay below .1 micron
but that is the sort of thing that OKAD with manual
layout was designed to handle.

Using very conservative estimates for speed the
numbers are stil astronomical.  The effect
of scale alone would allow an incredible number of
multiprocessing nodes to be connected on a 6" wafer.
I ran some conservative numbers for designs with
various amounts of RAM on chip and they all came
out in billions of Forth mips.  It would be an interesting
path to take. 

Heat has always been the problem for this sort of
thing, but according to what Chuck demonstrated
with test circuits is that the reason heat has
always been a problem is that people have not
been using the right equations to design for
hot transistors.  Armed with the right transistor
model that problem might be tamed.

They are a prototype, not a production facility.  But with
the size of our designs a single wafer of individual chips
would be a pretty big production run.  They mention the
user fees at the site but don't give any numbers.  I hope
it isn't one of those, if you have to ask you can't
afford it situations. But it is an interesting lead.

Jeff Fox