FLOATING POINT INSTRUCTION AND SELECTABLE COMPARISON ATTRIBUTES Russian patent published in 2020 - IPC G06F9/30 

FIELD: physics.

SUBSTANCE: invention relates to computer systems. Computer-implemented method of processing data for calculating floating-point numbers, comprising obtaining a command to be executed, commands for performing a comparison of a first value and a second value, as well as executing a command, wherein the instruction is a separate architecturally executed command which supports at least different functions for detecting a maximum or different functions for detecting a minimum, detection, based on the command control element, of the comparison function to be executed, selecting from several optional comparison capabilities for a comparison function of such an optional comparison capability, which corresponds to the first and second values to be compared, comparing a first value and a second value using an optional comparison capability to obtain a result, and the result is to be used when processing in a computing environment.

EFFECT: technical result is aimed at increasing the speed of operations with hardware in the environment of different programming languages, within their libraries by setting the comparison function for the first and second values.

12 cl, 33 dwg