CMOS 6502 timeline

[litwr]

I am curious to find when did 65C02 begin to be produced? https://en.wikipedia.org/wiki/WDC_65C02 gives 1978. This contradicts a bit the publication in AIM65 Interactive of 1980 about so called JMP (xxFF) bug. This publication at page 12 contains the phrase "There is a error in the JUMP INDIRECT instruction of ALL 6500 family CPU chips, no matter who they were made by." (http://archive.6502.org/publications/ai ... tive_2.pdf) So I can deduce that CMOS 6502 was not available in 1980 or it had the same JMP (abs) instruction as NMOS 6502. Am I right? Could anybody help with it? A lot of thanks in advance.

[BigEd]

Or they might have meant by 6500 family to exclude 65C00??

[Tor]

IIRC the AIM-65 were always produced with NMOS 6502, never with the CMOS 65C02. So I suppose the focus would be on the NMOS variant in the documentation.

[BigEd]

I see from August 82 CALL A.P.P.L.E that a CMOS 6502 was in debug, and that Rockwell would produce chips, and that MOS Technology would be a second source. This is a bit later than I would have guessed.

[hoglet]

This article reproduced in Atom Nieuws (a Dutch Acorn Atom Newsletter) also suggests samples were available in August 1982:

http://www.acornatom.nl/atom_nieuws/198 ... 828062.htm

This is also quite interesting (92MB PDF):
https://ia800201.us.archive.org/21/item ... _65C02.pdf

The data sheet is dated November 1982 and is preliminary.

Dave

[BigEd]

"Southwestern Data Systems Presents the wonderful All New 65C02"

I find it quite interesting as to who did what and when. WDC was founded in '78, apparently. Perhaps Bill Mensch created it so he could offer his design services. Perhaps he offered his services to Rockwell, an existing 6502 licensee, and perhaps he cut a deal to do the design for less money if his company could keep the design data to license elsewhere. I do recall some comment to the effect that MOS aka Commodore agreed to leave WDC alone in exchange for a free license - or something like that.

Also interesting that the 65c02 was not quite finalised but the '816 was already in the works. Feels to me that Rockwell wanted the bitwise instructions, but WDC decided to NOP them out to leave room for compatibility when the '816 came out.

Edit: see here
https://books.google.co.uk/books?id=JjI ... &q&f=false
where it says that WDC did the design, and initially licensed to Rockwell, GTE and NCR. Only Rockwell had the extra instructions.

[Tor]

The Apple Assembly Line (Bob Sander-Cederlof) December 1982 issue wrote "[65C02]. This is the enhanced CMOS version, soon to be available from GTE and Rockwell. [..] Rockwell is sampling the 65C02 now, and scheduled for production in Feburary [1983]."
The March '83 issue said "Well, I still haven't heard any more than I mentioned a couple of months ago."
The June '83 issue stated: "Don Lancaster just called to report that he has gotten his hands on samples of the GTE G65SC02 processor."

[ttlworks]

BigEd wrote:

Also interesting that the 65c02 was not quite finalised but the '816 was already in the works.

Interesting: now this probably could explain why the SBC instruction in decimal mode with non_BCD values gives different results for 65C02 and 65C816...

[litwr]

1983! Wow! It is a bit unexpected. Thanks for all the help. So WDC was producing NMOS 6502 up to 1983, and Rockwell, GTE, NCR too...

[BigEd]

I think this is a case of Wikipedia being in a muddle (check the Talk page.) I can believe WDC were started in 1978, but I can also believe they had no branded product until 1982 or 1983 - they called themselves a design center, and they made designs. (These days they describe themselves as a fabless chip company, but the term was, I think, not in use at the time.)

So, it seems WDC might well have originated all the CMOS 6502 designs, and eventually did use foundries to make WDC-branded chips. Did they make a 65C02 before their 65816? I'm not at all sure. It seems that CMD and VTI (aka VLSI) made '816 chips, and I wouldn't be surprised if their chips - licensed from WDC - came well before WDC's own.

The thing is, selling licenses for a good CPU design to well-known chip makers could be a more assured route to a good market share than selling your own-branded CPU design.

Ref: http://www.cpu-world.com/CPUs/65816/index.html

I would think WDC never had any NMOS product.

[litwr]

I've just read an interview with Bill Mensch - https://www.commodore.ca/commodore-hist ... he-brains/ - where he more accurately determined the date of 65C02. It was the spring of 1982. However Bill sometime exaggerates things. In this interview he claimed that the 65C02 was about 20% faster than the NMOS 6502 for programming languages - IMHO it is an exaggeration.even though he said (in his Oral History) that it was a claim by Steve Wozniak.

[L0uis.m]

I've always thought it was a pity that the first 6502 wasn't released in 1966