Subject: Re: FPGA VAX update
From: William Donzelli <aw288 at osfn.org>
Date: Sun, 06 Nov 2005 09:38:47 -0500 (EST)
To: "General Discussion: On-Topic and Off-Topic Posts" <cctalk at
classiccmp.org>
I posted that the EGO homebuilt comnputer was in 1985 and there
was another article on the TOY.
Well I found it. BYTE January 1987, Phil Koopman Microcoded vs Hardwired Control
The deisgn is a 16bit fairly minimal machine that could actually run real code
if built. The article does not fully discuss the the total machine but the control
logic is covered. The rest is almost obvious. Likely chip count for the entire
machine could be under 60 TTL peices if octal latches and Tristate gates plus the
74181s were used. If GALs were used the likely chip count could be smaller.
TOY itself was 16 bits, 4 bit instruction and 12 bit address, Single accumulator.
The ALU was 74181 so the basic arithmetic and logical instructions are native to
that chip and the remainder are jmpz (jump if zero) and load (put value in Acc)
and Store (save Acc at location). Oddly enough 4 no-ops, so theres room for
instruction set improvement.
A machine like this goes far to demystify how computers work.
An aside to all this is one of the annoying things when I was studying
computers early one (3++ decades ago) was "computer books" would endlessly
detail logic at the gate and flipflop level. Maybe discuss arithmetic and
sequential logic. They never quite cross the line to how these blocks form
computers. It wasn't until the PDP-8 handbooks that there was a connection
of the ideas of sequential logic controlling arithmetic and logical blocks.
Allison