Monday, October 28, 2019

The Team Behind the 6502

Team 6502 (via Hacker News):

When it was introduced in 1975 by MOS Technology of Valley Forge, Pennsylvania, the 8-bit microprocessor sold for a fraction of the cost of other microprocessors, causing rapid decreases across the entire computing industry. Featured in such seminal products as the Apple I and II, the Commodore PET, and the BBC Micro, as well as Atari and Nintendo game consoles, the 6502 microprocessor has been the brains inside toys, office machines, and medical devices too numerous to mention. As one of the most widely used microprocessor architectures of all time, the CMOS related form of the 65XX developed by the Western Design Center is still in production today, with an estimated six billion units so far produced.

While the story of Chuck Peddle, the visionary who conceived of the 6502, and that of design team member and founder of the Western Design Center, Bill Mensch, are widely recognized and recorded, the stories of the other MOS Technology engineers and employees who also worked on the 6502 and their contributions are not. This website seeks to change that.

Harry Bawcom:

In the 1970s much of what today is done by computer had to be done by hand. In the case of the 6502, once the design of the chip was completed by the team that worked on the chip’s architecture, that schematic was given to the design layout team. It was our job to create a topological layout from that schematic, a layout of all the transistors which made up the 11 or so glass reticles, also called "masks" that were then used to create the chip.


Design rule checking (the physical spacing between metal stripes: too close together and they will always short out in manufacture, making the chip non functional) and verifying schematic to layout was all manual. It was done by coloring a plot with colored pencils again and again as coloring a plot guided your eye to notice incorrect spacings or design rule violations.

Sydney Anne Holt:

We were given the logic drawings from Bill Mensch and Ray Hirt and etc, and turned them into the drawing you see in the picture from the Electrical Engineering Times article from 1975.

To do this, we drew them in pieces on big sheets of mylar that fit together like a puzzle. In order to do a careful logic to layout check we taped all the pieces together on the floor and crawled around on it to trace out the lines. The drawings were then digitized into layers so masks could be made from them.

I remember that once, one of the guys took off his shoes and was on the mylar checking when it was discovered his socks were damp and his toes were erasing the drawing as he moved along. Fortunately, it was caught very soon so the rework was minimal.

3 Comments RSS · Twitter

The newly developed Commander X16 (new developed Commodore-esque computer) uses this. See:

Nice find. I started programming on the 6502, writing machine code on paper and then "simulating" it by hand, testing if I got it right, after having found a book in the library about a 6502 board computer that have a 4x5 keyboard (for hex input) and a 7-segment display I could not even afford to buy.

Also, the best 8 bit CPU ever was the 6809, IMO. It was a pleasure to write (and read) its machine code.

The ugliest was the Z-80, followed the whole Intel range starting with the 8088. And look what we're all using now...

What I meant to say: The 6502 had a great design from a user's point of view. When I watched Michael Stein's presentation about the Apollo computer's design I realized the tricks they had to employ back then due to shortage of resources. Later, the Zylog and Intel CPUs seemed to come from the same designers, whereas the 6502 was something much cleaner. Or so I'd like to think. Probably not, but you learned to appreciate to program the 6502, 6809 and 68k CPUs when you also had to deal with the Zylog and Intel abominations :)

Leave a Comment