Path: utzoo!utgpu!news-server.csri.toronto.edu!qucdn!leek
Organization: Queen's University at Kingston
Date: Friday, 31 May 1991 16:20:01 EDT
From: <LEEK@QUCDN.QueensU.CA>
Message-ID: <91151.162001LEEK@QUCDN.QueensU.CA>
Newsgroups: comp.sys.amiga.graphics
Subject: Re: Idea for a graphics board
References: <1991May28.191019.13406@nntp-server.caltech.edu>
 <479@regina.uregina.ca> <1991May29.212315.27847@nntp-server.caltech.edu>

In article <1991May29.212315.27847@nntp-server.caltech.edu>,
nygardm@nntp-server.caltech.edu (Michael T. Nygard) says:
>
>I've been looking at some numbers on the board.  Here's what I came up with.
>
>Without using a CLUT (yuk.), assuming 80ns RAM, 1024x1024x24 resolution.
>
>1/(1024*1024*24*60) = .66227 ns/bit.
>(80 ns/cycle) / (.66227 ns/bit) = 120.8 bits/cycle
>
>That means a bus _at least_ 128 bits wide.  A 128 bit bus results in 5% of
>the memory cycles availible for the processor.

Hmmmm.  What about subdiving the 24 bit planes into smaller portions ?  eg
1, 2 or 4 bit planes per bus or better still - per board.  They'll require
a lot less bandwidth per bus so something like 32-bit or 64-bit should be
enough to allow for getting some CPU/graphics coprocessor cycles.  Some sort
of multiplexor scheme would allow the CPU to talk to one of these buses at
a time.  It would also be possible (not necessarily feasible) to allow the CPU
to access pixels on a word basis by accessing the same pixel location
on all the boards simultaneously. The multiplexor circuit is going to be
messy. :( Might be able to put it in a gate array. )

All of these bit planes would be combined to the D/A converter.  (Of course
they are required to sync up to the same DOT clock for generating the pixels.)
The individual boards are identical so the user can upgrade to more bit-planes
if they need it or have $$$.  It is very expandable in that aspect.  On the
hardware side, each of these boards can have its own graphic processors to
speed things up. (Most bit-plane operations can be run in parallel...)

>
>Other problems:
>I hadn't planned on this ever being a commercial product.  Getting the kind
>of printed circuit board this would require in single quantity would be very
>expensive.

My scheme allows you to build 1 bit plane and test it out.  The boards are
much smaller and probably easier to layout.  You would just have to order 24x
more boards when people start ordering from you.

>
>As for the cost estimate, it's hard to say.  Could be real high, maybe not.
>There are too many factors at this stage (early design stage).  No less than
>$1000 for a one-of-a-kind.  Now if we could put together a few people...
>
>This probably belongs in .hardware, but it started here, and I think the
>main interest is here.
>
>Free free to contact me if you have suggestions (or would like to help!)
>
>--
>--------
>Mike Nygard
>nygardm@coil.cco.caltech.edu

K. C. Lee
"Be careful of what you wish for.  You might get it."-Master Splinter,T.M.N.T.