Circuit Area Optimization Based on Approximate Computing and Majority Cover

In view of the problem of the minterm-based approximate computing techniques which are not suitable for area optimization of large circuits,a novel algorithm based on product terms and logic covers is presented.The proposed algorithm employs the product term based majority cover technique for approximate sub-functions searching and uses the disjoint sharp product operation for error rate computing.Unlike the minterm-based algorithms,the proposed algorithm can optimize large circuits efficiently by using product terms instead of minterms increasing rapidly in large circuits which lead the minterm-based methods unable to run efficiently or even out of work.The proposed algorithm has been implemented in C and tested under MCNC benchmarks.The experimental results show that the proposed algorithm can work fast even for large circuits.The number of inputs of the circuits has little effect on the speed of the algorithm.And the average area saving is 42.29%at the cost of the average error rate 3.67%.