I don’t know how I got into this rabbit hole but…
The wiki article on the CIA claims that its predecessor was the PIA. However, it would seem the VIA (6522) is more directly related, both in function and in products (the CIA replaced the VIA in Commodore products, IIRC). But I’m not really familiar with any of these.
Does anyone know the lineage of these products?
My guess would be, PIA is first. It’s possible the CIA and VIA both inherit from there? But here’s a tentative chronology:
PIA: MC6820 and later the MC6821 (seen in 1975 motorola applications manual)
VIA: 6522 and later 6256 (see 1977 preliminary info)
CIA: 6526 and later 8520 (see 1981 preliminary info)
The CIA is described here as an improved VIA
The CIA 6526 was not just a VIA with the shift register fixed. It included more elaborate timer handling, and a TOD clock, probably among other things.
I see all three of these have been tackled by the chip dissecting people
where we find a thread for each, containing die shots, floorplans, circuits, and notes:
Among the notes:
8521\6526 internally are completely different from the 6522, and to me it feels like the MOS designers simply had re_invented the wheel after Bill Mensch ™ had left MOS.
Also note:
We happen to know that the Rockwell 6522 had a completely different chip layout from the MOS 6522, so it would be interesting to check whether 6522 chips from other manufacturers really are cycle compatible to the MOS 6522 or not.
How does the time frame compare with 6800 PIA?
The January 1976 MCS6500 Microprocessor Family Hardware Manual from MOS Technology (available on BitSavers) includes details on the 6520 PIA and 6530 RRIOT.
The Nov. 1977 preliminary datasheet for the VIA (6522) refers directly to the PIA (6520):
MCS6522 VERSATILE INTERFACE ADAPTER
The MCS6522 Versatile Interface Adapter (VIA) provides all of the capability of the MCS6520. In addition, this device contains a pair of very powerful interval timers, a serial-to-parallel/parallel-to-serial shift register and input data latching on the peripheral ports. Expanded handshaking capability allows control of bi-directional data transfers between VIA’s in multiple processor systems.
So we may think of the VIA as an extended PIA. (It seems, while the PIA provides what may be needed in terms of internal interfaces, the VIA is meant to also provide what may be needed for inter-processor communications.)
In contrast to this, the CIA seems to be more like a fresh take on the problem. (May we interpret this also as breaking loose from the Motorola heritage?)
William D. Mensch, in remarks at VCF East earlier today, remarked on the original 6501/6502 being envisioned as the first chip of a two chip solution. The first was the microprocesso, of course, and the second the 6530 RRIOT - RAM, ROM, I/O, Timer. The intention was to compete with a cheap two chip 4040 offerring from Intel for the embedded market - and they initially gave little consideration to general purpose computing, except perhaps by Chuck Peddle.
Thanks for the instant reporting!
Chuck Peddle remarked that one of the first applications, they had in mind, was electronic cash registers (which is probably why the 65xx features BCD math). I can totally see how the PIA and VIA would fit in: you have to control local peripherals, like the keyboard, a display, etc., but then, from time to time, you have also to communicate the ledger exterally (which may require some additional abilities).
Well I think the idea of “general purpose computing” was not yet a thing. I seem to recall reading similar comments from multiple teams being baffled by all the people buying the various CPUs to use in homebrew computers.
One of the forgotten greats is the F8. Like the system you describe using a CPU and RIOT, the F8 was everything you needed in a two-chip package. RAM, ROM, CPU, ports, etc. Good price too.
Great idea until you remember, ROM was mask programmable only.
It was the EPROM that made BASIC and GAME cartridges possible,
Beg to differ: at any significant volume, manufacturers would use mask ROMs. They are much cheaper.
Back in 1983 I got a quote from Motorola and the crossover point where the 68701 (EPROM) became more expensive than the 6801 (ROM) plus the cost of creating the custom mask was just 100 units. I had expected a far larger number. I don’t know if pure EPROM / ROMs chips would have a similar ratio, but I expect so.