AMSYS29 floppies

Roger Holmes roger.holmes at
Mon Nov 22 04:05:10 CST 2010

> From: Al Kossow <aek at>
> A while back, part of an AMD 2900 development
> system showed up on eBay. My copies of the software
> surfaced today, so the .imd images are up under
> They are CP/M 1.4, as I recall.

Thank you. Might be extremely useful some time if I get around to using my similar system. On the other hand I might sell it, in which case having the image available on BitSavers has got to make the system more viable for someone. Just too busy with my 1301 in the summer and with my 1301 gate level simulator in the winter at the moment. Perhaps I should do a 2900 based computer with the 1301 architecture. Maybe an ICL 1900 and/or an Elliott 920.. The 920 ATC actually was 2900 based but I doubt the microcode has survived, though the early 920s would be simple as they only had 16 instructions - before they ballooned for CISC and then were reduced again for RISC. I suppose with a writable control store I could do anything from a Babbage Analytical engine or a Turing machine to the latest systems, though maybe not at full speed. Some of the more interesting machines (to me) like the ICT 1302 would be impossible because alas no full documentation of the instruction set survives. I have marketing information on it which amazingly actually lists some of the extra instructions but its far from complete and omits most of the actual function codes. Maybe I should ask the designer of the 1301 if he knows what his successors did.

Enough rambling, at least it was on topic.

Roger Holmes.
Technical Director, Microspot Ltd
Developers of CAD and Graphic software for the Apple Macintosh

More information about the cctech mailing list