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

Chuck was right


Anyone see this annoucement in EDN magazine:

        "Integrated Silicon Systems (ISS) has developed 
        [Psi]Time, a delay calculator that addresses the 
        problems of calculating delays in interconnect-
        dominated submicron chips...determining the RC
        network's 'effective' capacitance..."  
        EDN July 16, 1995.
        

Doesn't this sound like Chuck Moore's OKAD system he is
using on F21, P32, and P8?  (See the July/August 1995 
Forth Dimentions for some extra info on OKAD.)  Jeff Fox 
reported that Chuck

        "has been thinking about sub .1 micron technology.  
        It will be a big problem for the industry, because 
        at these scales the interconnect delay becomes more 
        signifigant than gate delay...Chuck's answer is what
        he does in OKAD, manual place and route.  Chuck says
        the industry will need some very advanced AI software 
        or they will need to get engineers involved in the 
        designs to solve these problems."

Well, you can get part of such software from ISS, but it 
costs **$25,000** and runs on a (presumably RISC) workstation.  
And from what I gathered from the announcement, that just buys 
you the delay information.  You still don't get any "AI" to help 
with silicon routing.

Maybe Chuck could shrink-wrap his OKAD simulation software 
and lowball the competition at $1,000 per package.  He could 
provide a way to read in silicon layouts in standard file 
formats so the EEs would be comfortable.  Maybe disable the 
manual place and route (editing) features until they buy an 
additional "upgrade."  You know they would be dying to tweak 
the layout manually once the package shows them delay 
bottlenecks!  Might need to make the OK inferface coexist with 
MS-Windows so the EEs' managers will pay for it.  (It can 
probably run in a DOS box if it gets its memory from DPMI
services).

I'd like to see Chuck make some money from his inventions, even
if most of the computer industry ignores his delightful MISC 
chips.  I see some other interesting business opportunities here.  
How does "Sub-Micron Silicon Tuning Services" sound?  A company 
could hire Computer Cowboys or an offshoot to hand tune a mostly 
finished design.  I know that approach won't yield optimal designs, 
but I can imagine that a company would spend BIG BUCKS to fix a chip 
that just misses its timing specs a little.  Especially near the end 
of a costly development/prototyping phase.

Physics majors unite!
--
Mike Losh                       Standard disclaimers apply.
B.A. Physics, Grinnell College