A project log for Suite-16

Suite-16 is a 16-bit cpu built entirely from TTL. It is a personal exploration of how hardware and software interact.

monsonitemonsonite 10/20/2019 at 15:070 Comments

Over the last few weeks, I have realised that the inspiration for my Suite-16 project most definitely has it's origins firmly rooted in the mid-1960s, a bit like me really......

I have illustrated some of my posts with hardware details of the PDP-8, PDP-11, Data General Nova and the IBM 1130 - all of which appeared in the period 1965 to 1970.

This weekend I am trying to chart out the direction this project will head in - based on my experience and skillset.

It's probably a given that there will never be a C-compiler for Suite-16. It's also unlikely that there will ever be a BASIC implementation, unless I follow the same course as the Gigatron TTL computer and first emulate a 6502 or Z80 processor.

So where does that leave this project without an obvious route to a high-level language?   (ShitCreek && !Paddle) ?

Given that the technology of this project is currently hovering somewhere between 1965 and 1975, a contemporary solution to this problem would be Charles H. Moore's Forth Language.

First - some ancient history

Back in the Fall of 1981, a school friend told me about a new language that he had read about in Byte Magazine.  Neil was the archetypal Geek who had told me about Z80 assembly language a couple of years earlier.  At that time I lived on the Isle of Man - a small Island in the middle of the Irish Sea, which, technology-wise,  was clearly a few years behind the times.  Neil told me about Chuck Moore and Radio telescopes and a whole lot of geeky stuff I didn't understand.  It almost seemed heretical that he was telling me about another language other than BASIC - because clearly that was the only thing I had encountered on the home computers of the day.

Some 2 years later I bought a ZX Forth ROM for my ZX81, and then in a surplus sale, a Jupiter ACE.  I dabbled in Forth but never really got off the starting grid.

After University, about 1987, and working for BBC Research Department, I bought one of Chuck Moore's Novix NC4016 Forth cpu development boards. This was the fastest cpu IC when it appeared in 1985 - and could outpace an 80286 manifold.

Despite my exposure to Forth at some key points in its evolution, I never seemed to practice the art to the point of becoming competent. 

So after almost 40 years my Forth skills are about as weak as my verilog and C skills - and part of the purpose of this project is to try to make amends and strengthen some of my former weaknesses.

And So Forth

Forth is a problem solving tool and integrated programming environment. It is self hosting, compact and versatile. It's also probably the most obscure programming language that you have ever encountered.

Forth evolved in the late 1950s and throughout the 1960s, and until 1970 was the exclusive programming environment of Chuck Moore.

It was commercialised by Forth Inc. throughout the 1970s and became mainstream by 1980, as featured in the August 1980 edition of Byte Magazine. 

Forth is based on a 16-bit virtual cpu model that can be implemented on almost any minicomputer, microprocessor or microcontroller.  It is ideal for standalone projects where memory and I/O resources are limited.

Forth may be implemented in 8K bytes of source code, or where space is limited a working kernel can be crafted into fewer than 1K bytes.

Moving Forth

It is my intention to use Forth to help solve my immediate computing problem - that is the design, construction and commissioning of a working 16-bit computer system.

As it's Sunday, I will keep this log brief.  Expect future updates in due course.