Hierarchical approach to module placement using neural networks

A neural network model for solving optimization problems proposed by J. J. Hopfield et al. has yielded good solutions of the traveling salesman problem. The present paper describes a method of solving module placement problems in automatic LSI device layout, in which a neural network model is used to optimize a set quadrisection procedure by minimizing its objective function. The solutions obtained for this method are comparable to or better than those obtained by the GFDR method, an iterative-improvement method of solving the module placement problem.