characteristic underflow

characteristic underflow

[‚kar·ik·tə′ris·tik ′ən·dər‚flō] (computer science) An error condition encountered when the characteristic of a floating point number is smaller than the smallest limit imposed by the hardware manufacturer.