Yes, it seems PALM did have a few evolutions, which just makes me curious if there were even earlier editions than this one from 1972. But even if so - then like the 4004, we're struggling to find evidence of actual products that made use of them. Wasn't the 4004 used in some cash registers, street lights, or some weighing machines? (I don't have any specific references, just recollections from past reading)
My suspicion is that something like the PALM was used in the large (later model) IBM 9-track tape systems (forget the model numbers offhand, but in their manuals they describe a full instruction set) or "industrial systems" like that. I think at that time (1972) PALM was also lacking a SHIFT or ROTATE code - since in part of the Joe George tech manual, they mention having to implement this in the SCAMP prototype themselves (a hardware solution workaround to a missing processor capability) until that instruction got added later (before the IBM 5100 release). [ specifically it is mentioned by Pat Smith in an entry from February 1973, who had come up with the workaround; this SHIFT I think was essential to some keyboard integration work ] Note that there is also evidence that as many as 400 early IBM 5100's were actually made in 1974 (based on an early bulk order of keyboards of that quantity, and a sales projection graph that includes a column for 1974) -- those early ones most likely were all APL only (since that was all that was yet available as it carried over from the SCAMP). Very nice images of the Philips tape. Yes, in the last few pages of that System/7 document, it has the 1972 article from Eletronic News about it - IBM was pretty proud of that, and it's the earliest example I've come across of using audio cassette tapes for digital data storage. So right around '71/'72 was the origin of that, as far as I know (at least as far as used in commercial products). Steve On Mon, Nov 27, 2023 at 2:56 AM Christian Corti via cctalk < cctalk@classiccmp.org> wrote: > > Well, just to throw this into the conversation: > > > > Over this past summer, I was studying the SCAMP ( > > https://voidstar.blog/scamp-a-review-50-years-later/ ) > > > > In that collection I came across a very early printing of the PALM > > instruction set, with the cover page dated March 21, 1972 of the > printing, > > and on the next page a date of March 16, 1972 of the document number. My > > photos of that document is here: > > > https://github.com/voidstar78/SCAMP/blob/main/IBM_SCAMP_PALM_InstructionSet_March1972.pdf > > This seems to be an older revision than the photocopied document that > I have. The instruction set described in the '72 document is not the final > one. Some opcodes are missing or are not complete (like the JUMP > instruction). > A transcription of my photocopy is here: > > http://computermuseum.informatik.uni-stuttgart.de/dev/ibm_5110/technik/instr_set.html > > BTW voidstar also has a document called System/7 tape cassette attachment. > I do have the original IBM cassette recorder (a Philips EL 3302) with > cable and System/7 diagnostics cassettes ;-) This was the tape recorder > used with the SCAMP. > Pictures can be found here: > http://computermuseum.informatik.uni-stuttgart.de/pics/ibm/s7 > > > Now, of course an argument is then is PALM a microprocessor? Perhaps not > > by todays standards and expectations, as it is a series of about 14 > > "Dutchess" chips, which is claimed to consist of MOSFET. I'm not enough > of > > I'd say yes. It's not a single-chip processor, but the i8008 wasn't either > (it couldn't work without support chips. > > > Christian >
