Japanese boffins try 'token passing' to scale quantum calculations Apart from actually performing computations, one of the most difficult quantum computing challenges is getting qubits to scale. A Japanese team has published what it believes is a solution to the problem of scale. Quantum gates are complex creatures with many more components than their classical equivalents, so instead of …
My increasingly vague and faulty memory (it's alcohol-erasable and it's been through too many cycles) dredges up early computer designs where the ALU was one bit wide and operated sequentially. Data was supplied by accoustic delay lines implemented using tubes full of mercury.
Something like that, anyway. The name "Ferranti" also seems to be ringing bells in my dodgy memory.
A quick google turns up such gems as UNIVAC 1, LEO 1, Highgate Wood Telephone Exchange and "various Ferranti machines." Nothing explicitly mentions single-bit ALU design, but more vague flutterings of memory seem to imply I was told this at a university lecture, many decades in the past.
Vague and incomplete my memories may be, but I'm fairly sure there were 1-bit ALUs in the early days. They may run 8 times more slowly than 8-bit ALUs, but back when the circuitry was really expensive, it was a good trade-off. For starters, you don't need to worry about carry look-ahead circuitry. :)
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
