An efficient algorithm for microword length minimization

The problem of microword length minimization is crucial to the synthesis of microprogrammed controllers in digital systems. The authors formulate the problem as a graph partitioning problem. They employ a local search approach to further reduce the microword length. The algorithm is capable of finding fast and near-optimal solutions for very large size microcodes, efficiently. The algorithm was tested with practical microcodes and with large size examples generated from random graphs. The experimental results are compared with those of other methods.<<ETX>>