This document compares the performance of three multiplier architectures - array, radix-2 Booth, and radix-4 Booth multipliers. It first describes how each multiplier is implemented and its working. It then discusses simulating the multipliers in Xilinx and analyzing the results. The radix-2 Booth multiplier is found to have the best performance in terms of delay and power consumption by reducing the number of partial products. The document concludes the radix-2 Booth multiplier is best for high-performance applications due to its optimized speed and lower power usage.