Relay computerswhy so few 17
But obviously vacuum tube logic *could* have handled all of that, but wasn't used (because it was too expensive). Likewise it is also true that solid state was able to handle the line interface long before relays were replaced with an electronic network fabric.
The actual problem with conversion from relays to totally electronic switchin *was* the logic! Until an all digital switching fabric was available in the mid-1970's, telephone switches were still mechanical relays. They were common control, and that was provided by an all electronic computer, but it still came down to a mechanical relay!
Relay computerswhy so few 18
You might well have designed a PBX, but nobody designed an all digital switching fabric suitable for a 10,000 line CO switch until somewhere around 1975-6. There is...
The significance of that turned out to be great too. WECO continued to design switches by adapting new common control systems to the old all relay switching fabric concepts. Northern Telecom Inc did something different, and designed a common control system "as if" there existed a digitial switching fabric, and then designed a relay module to plug into it to emulate a digital switching fabric that wasn't yet available.
When the day came that the technology was available, NTI designed an all electronic digital switching module and plugged into their existing control system, and instantly had an all digital telephone switch. WECO had to begin the design of a whole new switching system to take advantage of the digital switching fabric, and by the time it was ready to go NTI had taken literally 40% of the market away from them.
Supposedly the bright idea at NTI to do it that way actually came from the CEO himself.
Relay computerswhy so few 20
Tim Shoppa I must disagree. After WW II there was a tremendous need for computing horsepower. IBM sold out a prototype electronic calculator (Type 603, part of its punched card...