| 摘要 |
PURPOSE:To easily process scaling, etc., by suppressing the generation of the number of floating points when the length of a detected exponential part is longer than that of the designated exponential part, and notifying the fact to the outside. CONSTITUTION:After the completion of a desired calculation by means of an arithmetic part 5, the exponent as a result is read from a exponential register group 2, an exponent 103 is inputted to an exponential part length detector 6, and an exponential part length 106 is detected. Further, the length is compared with an allowable value 111 set in an exponential part length designating register 8 by a comparator 7, and when the detected exponential part length 106 is longer than the set value, the generation of the number of the floating decimal points is suppressed, and the fact is notified to the outside by an interrupting signal 109. Thus, the calculation at desired accuracy can be executed, and since the detected exponential part length and the code of the exponent can be read, the scaling, etc., can be easily processed. |
