Path: utzoo!censor!geac!becker!hybrid!scifi!bywater!uunet!dptechno!dave
From: dave@dptechno.UUCP (Dave Lee)
Newsgroups: comp.theory.cell-automata
Subject: Re: John Conway's "Life"/Pseudo-Neural hardware implementation
Message-ID: <606@dptechno.UUCP>
Date: 7 Nov 90 20:08:06 GMT
References: <1990Nov1.185409.25802@bradley2.bradley.edu> <10440@milton.u.washington.edu> <2371@oucsace.cs.OHIOU.EDU>
Reply-To: dave@dptechno.UUCP (Dave Lee)
Distribution: usa
Organization: D.P. Technology Corp. Camarillo California
Lines: 27

Just to add to this discussion, I built a 16x16 led grid driven by 
a Z80 in 1985.  It was primarially for animation, but I coded in 
a life game for fun.  It used a homemade optical pen made out
of a photo transistor imbeded in the empty shell of a bic pen.

Anyway, as for a hardware solution being faster than a software one,
well it may be, but how fast can you watch life ?
I cant remember the details, but my system only took about 5 instructions
per cell to compute, and running at a clock of 4Mhz, I had to put delays
in all over just to make it watchable.  

Having each cell do its own computing is certianly elegent, fun, and
probably useful for "serious" computing, but for watching a life
game in real-time, a simple cpu with a little memory can certianly
keep up with any desirable display speed. (IMHO of course).
The chip count (and cost of course) of a simple cpu based system 
doesnt grow much with the number of cells.  Especially if you use
multiplexed display hardware (one transister and 1 bit of latch 
per row and per column)  
No matter how you do it, it's a pain to wire all those led's !!!!!
More reciently, I've seen LCD grids availible (like those in hand held
games) that would make the display hardware a snap.


-- 
Dave Lee
uunet!dptechno!dave