> On Jul 5, 2025, at 10:14 PM, Gavin Scott via cctalk <[email protected]> > wrote: > > On Sat, Jul 5, 2025 at 2:40 PM Paul Koning via cctalk > <[email protected]> wrote: >> I assume those were all done by hand; it's not obvious how a robot could do >> that in the early 1960s, unlike wire wrap backplanes. > > Nice page on the hand-assembly of early Cray systems with cool pictures: > > https://cray-history.net/2022/08/28/sonjas-story-about-cray-1-fabrication-and-assembly/
Wow, very nice! The CDC 6000 backplanes look quite similar but they are much bigger. https://upload.wikimedia.org/wikipedia/commons/c/c4/CDC_6600.jc.jpg Speaking of core memory: conventional RAM is easy to wire (with steady hands) because the cores can be set in a holding jig and the wires then are threaded straight through. But there are core ROMs, and in those the data patterns are defined by wires that either go through or skip around a given core. And you have a whole bunch of wires (32 or more) per core. So you end up with a row of cores and a big bundle of wires randomly zigzagging in and out of the cores. The name "core rope memory" came from what that looks like. Occasionally such ROMs use cores that are merely transformer cores, for example in an early Wang calculator. But the earlier designs used memory type cores, essentially acting as logic elements. That is the case both for the best known example, the Apollo spacecraft on board computer, and for the Dutch Electrologica X-1 (which uses a different structure with faster data access timing). In fact, conventional RAM core memory also uses the cores as logic elements, they are not simply bit storage devices. The point is that each core essentially performs an AND operation on the wire current of the wires running through it, with the read operation defined as Xsel and Ysel, and the write as Xsel and Ysel and not Inhibit. The magnetic properties of the cores make this possible. Some early logic circuits were magnetic ones, built out of memory cores. I remember a device documented in the ham radio magazine QST (mid 1960s) that used magnetic core shift register. And Ken Olsen (founder of DEC) got his MS from MIT via a thesis on magnetic core logic devices. (Oh yes, and his name appears as contributor to the Apollo computer ROM design.) paul
