专利内容由知识产权出版社提供
专利名称：NUMERICAL EXPRESSION CONVERTER AND
VECTOR PROCESSOR USING THE SAME
发明人：HATTA, Koichi申请号：EP92906919.3申请日：19920318公开号：EP0530372B1公开日：19980527
摘要：In a device for converting a floating-point expression into fixed-pointexpression, a decimal alignment means (12) shifts the digits of the mantissa (M) of anumber to form a fixed point number depending on the exponent (E). When the sign part(S) of a floating-point number is negative, the integer part of the shifted number isinverted by a bit inversion means (31). At the time of shifting, the value of a carry digit isdetermined by a carry determining means (32) based on the data that indicates whetherthere are digits on the right side of the decimal point and a truncate direction instructiongiven from an external unit. Thereafter, the value output from the bit inversion means(31) and the carried value determined by the means (32) are added through an addermeans (33) which then outputs the sum in the fixed point system. The above constitutionenables the operation speed to be increased and the circuit constitution to be simplified.Further, the above numerical expression converter can be adapted to a vector processorunit that calculates vector data.
申请人：FUJITSU LTD
地址：JP
国籍：JP
代理机构：Stebbing, Timothy Charles
更多信息请下载全文后查看