Early core memories were woven by hand, but IBM rapidly automated the process. (Since most computers from the 1950s to early 1970s used core memory, there was a lot of demand.) However, IBM later found that it was cheaper to have the memories assembled by hand in Asia. For detailed information on core memory, see the book "Memories That Shaped an Industry".
Core rope is different from core memory and much rarer. Core rope is essentially ROM, using much larger cores with wires going through or around a core, storing 192 bits per core. Core ropes were hand-woven (with machine guidance) for the Apollo Guidance Computer.
December/January 1987 I was doing a vacation EE internship in a power station in Australia. Some of the Hitachi mini computers still used core RAM. This was in an all Hitachi Heavy Industries turnkey coal-fired power station commissioned ca. 1985. Pretty sure they had a reference design from boilers and turbines right down to the hardware and software level and kind of cookie cutter stamped out power stations from it. The Hitachi engineering attitude was obviously "If it works, keep doing it the same way for as long as possible". I was told that for some software (firmware?) updates, they'd simply ship out a new core RAM module -- It's non-volatile after all.
Jacquard loom!
Yes, and early core memory was also woven by hand. I am not sure if this was just for core rope memory, or if it was more widespread than that.
Early core memories were woven by hand, but IBM rapidly automated the process. (Since most computers from the 1950s to early 1970s used core memory, there was a lot of demand.) However, IBM later found that it was cheaper to have the memories assembled by hand in Asia. For detailed information on core memory, see the book "Memories That Shaped an Industry".
Core rope is different from core memory and much rarer. Core rope is essentially ROM, using much larger cores with wires going through or around a core, storing 192 bits per core. Core ropes were hand-woven (with machine guidance) for the Apollo Guidance Computer.
December/January 1987 I was doing a vacation EE internship in a power station in Australia. Some of the Hitachi mini computers still used core RAM. This was in an all Hitachi Heavy Industries turnkey coal-fired power station commissioned ca. 1985. Pretty sure they had a reference design from boilers and turbines right down to the hardware and software level and kind of cookie cutter stamped out power stations from it. The Hitachi engineering attitude was obviously "If it works, keep doing it the same way for as long as possible". I was told that for some software (firmware?) updates, they'd simply ship out a new core RAM module -- It's non-volatile after all.
[dead]