Dynamic Weighting Monte Carlo for Constrained Floorplan Designs in Mixed Signal Application.pdfVIP

Dynamic Weighting Monte Carlo for Constrained Floorplan Designs in Mixed Signal Application Jason Cong, Tianming Kong, Dongmin Xu VLSI CAD LAB., Computer Science Department University of California, Los Angeles, CA 90095
cong, kongtm, dongmin @ Faming Liang, Jun S. Liu, Wing Hung Wong Department of Statistics University of California, Los Angeles, CA 90095
fmliang, junliu, whwong @ Abstract Simulated annealing has been one of the most popular stochastic optimization methods used in the VLSI CAD field in the past two decades for handling NP-hard optimization problems. Recently, a new Monte Carlo and optimization method, named dynamic weighting Monte Carlo [WL97], has been introduced and successfully applied to the traveling salesman problem, neural network training [WL97], and spin-glasses simulation [LW98]. In this paper, we have successfully applied dynamic weighting Monte Carlo algorithm to the constrained floorplan design with consideration of both area and wirelength minimization. Our application scenario is the constrained floorplan design for mixed signal MCMs, where we need to place all the analog modules together in groups so that they can share common power and ground planes, which are separate from those used by the digital modules. Our experiments indicate that the dynamic weighting Monte Carlo algorithm is very effective for constrained floorplan optimization. It outperforms the simulated annealing for a real mixed signal MCM design by   in wirelength, while gets slight area improvement. This is the first work adopting the dynamic weig