METHOD FOR LOADING DATA AND METHOD FOR STORING THE SAME AND METHOD FOR LOADING DATA WORD AND METHOD FOR STORING THE SAME AND METHOD FOR COMPARING FLOATING POINT

摘要

PROBLEM TO BE SOLVED: To efficiently use a 64 bit architecture especially for 32 bit and 64 bit data values in loading and storing data, and to provide a floating point instruction set for further efficiently interfacing with a CPU in comparing floating points. SOLUTION: The elements of a processor and a co-processor are constituted so that a floating point instruction set architecture can be executed. Then, double precision words and/or plural pairs of single precision words are loaded and stored between a floating point co-processor and a memory, and a Boolean result obtained by comparing two floating points in a floating point unit 265 is stored in an integer register in a CPU 201.