3 ms·
I accidentally discovered a year or so ago that Thomas L. Dimond (working at Bell) invented core rope memory (which the wikipedia page says was "first used in t
by zkms 9y ago
I accidentally discovered a year or so ago that Thomas L. Dimond (working at Bell) invented core rope memory (which the wikipedia page says was "first used in the 1960s") in the 1940s. His memory got used in phone switches as a fast and reliable read-only lookup table. Here is a paper about the "Dimond ring translator": http://etler.com/docs/Crossbar/articles/30-AMATranslator.pdf http://etler.com/docs/Crossbar/articles/30-AMATranslator.pdf
- Animats 9y ago"Writing" was done by manually stringing wires though those big rings. Those could be changed, but it was a big headache. It's like classic reverse DNS. There was already a mapping from A to B (phone number to outgoing wires) and a mapping from B to A was needed for billing purposes. But B to A info couldn't be obtained from the switch fabric. A physically separate B to A mapping had to be built and maintained in sync. There was nothing which inherently made the two match. That was all done by hand. The early history of computing was a struggle to find a usable memory device. Relays were very bulky as memory devices, and none of the relay computers had much memory.