engleski

On the possibilities of the BCD Code Application in the Floating-Point Arithmetic Algorithms

This paper deals with the possibilities of the BCD code application in the software realizations of the Floating-Point Arithmetic Algorithms using DAA instruction. Applying BCD and/or binary code to the exponent and/or mantissa representation of the Floating-Point Number, four different Floating-Point Arithmetic variants have been realized for the microcomputer Intel 8080. Essential parameters, like roundoff and truncation errors, algorithms execution times, dynamic range, program length, input/output conversion times, etc., are measured, compared and discussed in detail.

Floating point; Arithmetic; Algorithm; Truncation error; Roundoff error

nije evidentirano