On a drum-based computer, the execution speed of programs depends greatly on the sequence of instructions on the drum. I don't know the coding for the Univac II, but on the IBM 650 - the first computer I write code for - each instruction had two operands. The first was the address of the data for that instruction, the second was the address ofthe next instruction.
The easiest way to write code was to put the instructions in sequential order - but since the drum spun past two or three memory slots during the typical instruction, that resulted in nearly the slowest execution time, since the drum would have to spin an entire time around before it got to the next instruction.
IBM provided an assembly program called SOAP -the "Symbolic Optimum Assembly Program" - which optimized the location of instructions. As a freshman at Cornell in the spring of 1958, I convinced the Cornell COmputing Center to adopt SOAP and wired the board for it.
![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
Re the circular slide rule
Date: 2012-09-29 09:20 pm (UTC)The easiest way to write code was to put the instructions in sequential order - but since the drum spun past two or three memory slots during the typical instruction, that resulted in nearly the slowest execution time, since the drum would have to spin an entire time around before it got to the next instruction.
IBM provided an assembly program called SOAP -the "Symbolic Optimum Assembly Program" - which optimized the location of instructions. As a freshman at Cornell in the spring of 1958, I convinced the Cornell COmputing Center to adopt SOAP and wired the board for it.
Dad