Processor having efficient function estimate instructions

A computer-implemented method comprising:

• executing a first instruction on a first portion of a numerical operand to obtain a first instruction result that is a floating point number having a mantissa, a sign bit, and exponent bits, wherein the first instruction corresponds to a function to be estimated, wherein said mantissa contains a first bit field and a second bit field, wherein the first bit field contains the most significant bits of said mantissa and the first bit field represents a base value in a linear interpolation function corresponding to the function to be estimated, and wherein the second bit field contains the least significant bits of said mantissa and the second bit field represents a slope value in the linear interpolation function;
• and
• executing a second instruction on a second portion of said numerical operand and the first instruction result, wherein the second instruction evaluates said linear interpolation function, as defined by the first instruction result, at said numerical operand to obtain a floating-point estimate of the function to be estimated.