Design and Implementation for a Integer-Float ALU
-
Abstract
For solve the issue that commont floating pointe ALU comsumes large area and srtucture loosely, a floating point and integer mixed operator M-ALU are designed and implemented. The algorithm based on the base-4 algorithm Wallace tree structure, and try a new order code alignment method, merging integer arithmetic and floating point arithmetic processing logic. In the three-stage pipeline structure can be accurately completed single-precision floating-point, extended precision floating-point and integer basic operations. Based on the synopsys provided by the Design Compiler synthesis tool in the SMIC 65nm process library to complete the synthesis, to 500 MHz frequency.
-
-