An arithmetic underflow is the result of a calculation that is too small to be expressed properly. For example, in floating point, a negative exponent can be generated that is too large (too small a ...