Population-based FPGA solution to Mastermind game

Abstract

We present a new parallel, adaptable algorithm, which plays Mastermind game, and its FPGA implementation. The proposed algorithm is a cross between Shapiro's, Knuth's, and Kooi's algorithms, has low-computational complexity but still offers competitive game results. The FPGA design part required subtle architectural decisions and trade-off between area, interconnect, and clock speed. An FPGA implementation provides a cheap and low-power engine for a console game while offering speeds even greater than a high-end PC. The adaptability of the algorithm lets us easily move the algorithm to bigger FPGAs if greater speeds and/or wider versions of the game are needed. © 2006 IEEE. © 2008 Elsevier B.V., All rights reserved.