In the conventional iterative sum-of-products circuit there is a well-defined transition from the simply pipelined summation tree to a purely iterative final stage. As has proved to be the case for the partial product reduction trees in fast parallel multipliers, better results are obtained when the simple structural paradigm is abandoned. We follow the parallel multiplier developments by limiting carry propagation distance and extend them by distributing the iterative (feedback) elements throughout the tree.

Mon Dec 11 17:02:42 CST 2000