4 ms·

Computation was extremely expensive back then. Was it typically financially beneficial?

by leoh 7y ago

blihp 7y ago
In general, definitely. But it was only done for critical/high value tasks that either weren't feasible to do any other way or the cost savings from labor elimination made it cost effective. As recently as the 80's businesses typically weren't frivolous with computer purchases... that changed in the mid-to-late 90's.
pjc50 7y ago
The first commercial applications were for basic but critical accounting; payroll and end-of-period reporting. Although the LEO page doesn't have much info on what programs were run: https://en.wikipedia.org/wiki/LEO_(computer) https://en.wikipedia.org/wiki/LEO_(computer) > One of its early tasks was the elaboration of daily orders which were phoned in every afternoon by the shops and used to calculate the overnight production requirements, assembly instructions, delivery schedules, invoices, costings, and management reports. This was the first instance of an integrated management information system
- leoc 7y ago
  Here's a demonstration of and business case for LEO from 1957: https://www.youtube.com/watch?v=-8K-xbx7jBM https://www.youtube.com/watch?v=-8K-xbx7jBM .
TheOtherHobbes 7y ago
Computation was so financially beneficial that performance could be trimmed by a couple of orders of magnitude, and people still found it beneficial as long as it was interactive. For any application that didn't require scientific supercomputing, a mid/late-70s 8-bit interactive command line microcomputer with 16K of memory and paper tape or cassette storage was far more financially beneficial than a 7090 or System360 - not just in terms of cost, but in terms of the ability to get useful output in a quick-enough time frame. Speeding up the development cycle and simplifying data entry was more valuable than raw processor speed. The metric that mattered was problem throughput, not clock speed. Some problem classes were cycle-bound, but many weren't - and still aren't.
- leoc 7y ago
  And business calculations at many companies were automated or semi-automated by tabulating machines before those were finally replaced by fully programmable computers. The IBM 1401 released in 1959 was designed to make inroads into that market as a drop-in substitute or upgrade. Earlier computers would probably not have been financially viable in many of those roles: https://www.youtube.com/watch?v=ZPpV8X91neQ https://www.youtube.com/watch?v=ZPpV8X91neQ

Computation was extremely expensive back then. Was it typically financially beneficial?

by leoh 7y ago

Computation was extremely expensive back then. Was it typically financially beneficial?

In general, definitely. But it was only done for critical/high value tasks that either weren't feasible to do any other way or the cost savings from labor elimination made it cost effective. As recently as the 80's businesses typically weren't frivolous with computer purchases... that changed in the mid-to-late 90's.

pjc50 7y ago

The first commercial applications were for basic but critical accounting; payroll and end-of-period reporting. Although the LEO page doesn't have much info on what programs were run: https://en.wikipedia.org/wiki/LEO_(computer) https://en.wikipedia.org/wiki/LEO_(computer) > One of its early tasks was the elaboration of daily orders which were phoned in every afternoon by the shops and used to calculate the overnight production requirements, assembly instructions, delivery schedules, invoices, costings, and management reports. This was the first instance of an integrated management information system

leoc 7y ago
Here's a demonstration of and business case for LEO from 1957: https://www.youtube.com/watch?v=-8K-xbx7jBM https://www.youtube.com/watch?v=-8K-xbx7jBM .

TheOtherHobbes 7y ago

Computation was so financially beneficial that performance could be trimmed by a couple of orders of magnitude, and people still found it beneficial as long as it was interactive. For any application that didn't require scientific supercomputing, a mid/late-70s 8-bit interactive command line microcomputer with 16K of memory and paper tape or cassette storage was far more financially beneficial than a 7090 or System360 - not just in terms of cost, but in terms of the ability to get useful output in a quick-enough time frame. Speeding up the development cycle and simplifying data entry was more valuable than raw processor speed. The metric that mattered was problem throughput, not clock speed. Some problem classes were cycle-bound, but many weren't - and still aren't.

leoc 7y ago
And business calculations at many companies were automated or semi-automated by tabulating machines before those were finally replaced by fully programmable computers. The IBM 1401 released in 1959 was designed to make inroads into that market as a drop-in substitute or upgrade. Earlier computers would probably not have been financially viable in many of those roles: https://www.youtube.com/watch?v=ZPpV8X91neQ https://www.youtube.com/watch?v=ZPpV8X91neQ