Fw: f21 (or better) Code

["mlosh" <mlosh@xxxxxxx>]

-----Original Message-----
From: Me <deah@home.net>
To: MISC <misc@pisa.rockefeller.edu>
Date: Thursday, June 01, 2000 4:38 PM
Subject: f21 (or better) Code


>Why not make the VHDL or whatever is used, available to download, so we
>could make f21's with our existing FPGA's and PAL's? They wouldn't be as
>fast, of course, but they would be faster than software emulation, and
allow
>for some degree of hardware prototyping. It would also allow us to better
>understand how the damn thing really works (and if it really does).
>
>Open Source the f21 project!!!!
>
>

If I may, this suggestion shows a lack of knowledge of how these chips have
been designed and prototyped. I don't want to sound harsh, but if you read
more of the transcripts and articles at the http://www.ultratechnology.com
website, you would probably not ask the questions in the way you did.

I am not a hardware designer, so I may not have some of the details right,
but let me paint a general picture of Chuck's development process for the
MuP21 and F21 chips...

Chuck has developed his own totally unique software to design silicon
circuits. He does not view the chip fundamentally as a network of logic
gates. With his tools, he designs directly in the layers of material of a
silicon chip. Chuck "draws" patterns into the various layers (metal,
insulator, doped semiconductor, etc.) with a graphical editor and the
intersections of those patterns form the active and passive elements of the
intended circuits. Chuck knows how to lay out patterns to form transistors,
and how to group transistors to form gates and larger functional units.
Working at such a low level, he has an opportunity to customize the layout
and composition of circuits, and trim out unnecessary elements. The result
is a highly optimized design that can run far faster than a
conventionally-designed chip, which has several levels of conservative
safety margins built in. However, there are no schematics of logic gates
anywhere in this process.

Chuck's is a "risky" approach that pushes customary boundaries, and I think
it has caused setbacks because the fabrication prototypes don't always
behave as expected. Chuck has refined his simulation tools (which also
calculate signals at the silicon layer level) to be more accurate. When
Chuck is ready to fabricate a prototype, his tools convert the design
patterns within the various layers into a file containing the list of
rectangles that the lithography masks must contain. The fabrication plant
makes the masks and produces the prototype chips with them.

So in summary, there is no VHDL code to share, nor any other
industry-standard logic designs. The best you can get is a standard-format
lithography pattern file. To do anything with Chuck's designs, you would
have to have his tools and chip data and learn how to use them. Chuck says
that it is not difficult to learn, but I am not sure all hardware designers
could handle it... it would seem to me you would need to develop a
sophisticated understanding of how signals inside silicon really behave and
you could not pretend that the circuits are perfect logic gates.

Another issue completely is that Chuck is doing silicon design work as his
living, and I assume that he and his employers/investors are seeking a
reasonable return on their investments of time and money. Those of us who
are watching along the sidelines may hope that more things open up, but I
respect their decisions. The current success of some open source
technologies like Linux may encourage them to open more, but they need some
major design wins if the technology is going to really take off.

Chuck has said that he plans to release a PC-compatible version of his Color
Forth software for free. I speculate that Chuck could make a fair amount of
money by publishing as a book a contrarian critique of the current hardware
and software industry... along the lines of Cliff Stoll's _Silicon Snake
Oil_ or Jef Raskin's _The Humane Interface_. The book could contain a
Color-Forth system diskette/CD-ROM with email and a web browser to show how
much can be done with a tiny amount of code. With a good editor and a big
publisher behind him, he could probably sell 100,000 books or more, and have
some money for further MISC hardware R&D. I don't know if he wants to do
something like this, however.

Back to the designs... it may be possible to abstract the designs so that
they could be implemented in a FPGA, but that would take a substantial
effort. It would probably be easier to just design one yourself, if you have
the knowledge and skills. If not, there are some people who have started
such designs. There are videos for a silicon design class by John Rible (I
think?) available form Ultra Technology. Others in this list may be able to
describe such designs and may be willing to share them.

Have fun in your research and projects!
Mike Losh