Abstract Comparison of two floating point numbers is performed using a unique algorithm, which in one embodiment is implemented very efficiently in hardware, resulting in a very significant improvement in the floating point comparison execution time. In order to perform a floating point comparison, the exponents of the two numbers to be compared are subtracted. In many cases, the exponents are not equal, and simply determining which exponent is larger, combined with the signs of the mantissas of each of the floating point numbers being compared, is sufficient to determine which of the two floating point numbers being compared is larger. Alternatively, if the exponents of the two floating point numbers being compared are equal, the mantissas may be subtracted without need for shifting mantissas in order to equalize the exponents. In this manner, the floating point comparison operation is greatly simplified and the time consuming multiple shift operation is avoided, thereby vastly reducing the amount of time required to perform the floating point comparison, and thereby increasing the overall speed of operation of the microprocessor.
Means for comparing a first and a second floating point number comprising: first means for providing an exponent difference signal indicating if the exponent of said first floating point number is greater than the exponent of said second floating point number; second means for providing an equal exponent signal indicating if the exponents of said first and second floating point numbers are equal; third means for providing a signal indicating if either or both of said first and second floating point numbers are equal to zero; and fourth means for providing an output signal indicating if said first floating point number is greater than said second floating point number, by selecting as said output signal: the most significant bit of said first floating point number when said exponent of said first floating point number is not greater than said exponent of said second floating point number, and neither of said first or second floating point numbers is equal to zero; the sign bit resulting from the subtraction of the mantissas when the exponents of said first and second floating point numbers are equal and there has not been an overflow when subtracting said first and second floating point numbers; and the inverse of the sign bit resulting from the subtraction of the mantissas when the exponents of said first and second floating point numbers are equal and there has been an overflow when subtracting said first and second floating point numbers.
Claim 2
Dependent Claim
241 words
Means for comparing a first and a second floating point number comprising: first means for providing an exponent difference signal indicating if the exponent of said first floating point number is greater than the exponent of said second floating point number; second means for providing an equal exponent signal indicating if the exponents of said first and second floating point numbers are equal; third means for providing a signal indicating if either or both of said first and second floating point numbers are equal to zero; and fourth means for providing an output signal indicating if said first floating point number is greater than said second floating point number, by selecting as said output signal: the most significant bit of said first floating point number when said exponent of said first floating point number is not greater than said exponent of said second floating point number, and neither of said first or second floating point numbers is equal to zero; the sign bit resulting from the subtraction of the mantissas when the exponents of said first and second floating point numbers are equal and there has not been an overflow when subtracting said first and second floating point numbers; and the inverse of the sign bit resulting from the subtraction of the mantissas when the exponents of said first and second floating point numbers are equal and there has been an overflow when subtracting said first and second floating point numbers.