My first job as engineer 1963 was in a company which made office machines. A 
desk full of relays and 'Strowger selectors' - the same things which they 
used in telephone exchange at that time. People there liked to show the 
'speed' by entering a little task like 99999 x 999 then go to lunch and when 
we returned it was still working on it tak tak tak tak tak tak rrrrrrrrrt 
tak tak tak....
We developed the new generation machines using core memory and blocking 
oscillators (rather than flip-flops, monoflops) because they worked with 
only one transistor. Transistors were expensive, around 5 to 10 $ a piece.

Lembit


> >I did think that initialy, but just imagine how many
>>relays you would need for a 32kbyte memory! I find it
>>quite amazing that the rest of the system uses "only"
>>415 relays given the surprising amount of functionality.
>
> Probably get away with 32k relays - if one used magnetic latching ones.
> Would have the advantage they would be like core memory, and retain the 
> info
> across power outages.
>
> The bigger hassle would come in the address logic, but even that could be
> do-able. If the memory cell coils are individually brought out with one 
> end
> used as a row address, and the other as a column address, then the ladder 
> to
> decode the address would come down to a manageable size - the limiting
> factor would be the number of contacts you could put on a single relay for
> the final stage or two of the ladder.
>
> -- 
> http://www.piclist.com PIC/SX FAQ & list archive
> View/change your membership options at
> http://mailman.mit.edu/mailman/listinfo/piclist
>
>
> -- 
> No virus found in this incoming message.
> Checked by AVG Free Edition.
> Version: 7.1.375 / Virus Database: 267.15.8/260 - Release Date: 14.02.2006
>
> 

-- 
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist