A Two's Complement Array Multiplier Using True Values of the Operands

A new algorithm for implementing the two's complement multiplication of an m × n bit number is described. By interpreting certain positive partial product bits as negative, a parallel array is developed which has the advantage of using only one type of adder cell. A comparison with the Pezaris and Baugh-Wooley arrays is presented, showing that the new array is as fast as the Pezaris array and uses less hardware than the Baugh-Wooley implementation.