Evolutionary techniques in physical design for FPGAs

This paper disscusses two studies of using evolutionary algorithms in physical design for FPGAs. The first study presents an adaptation of a genetic algorithm that optimises parametres of a linear delay model for Xilinx XC4000 FPGA and compares their performance to parameters optimised by hand. The second study showes implementation and performance of an adaptive technology mapping algorithm for XC4000 based on Wilsons XCS classifier system