Automatic Code Generation on a MOVE Processor Using Cartesian Genetic Programming

TY - JOUR

T1 - Automatic Code Generation on a MOVE Processor Using Cartesian Genetic Programming

AU - Walker, James Alfred

AU - Liu, Yang

AU - Tempesti, Gianluca

AU - Tyrrell, Andy M.

PY - 2010/9

Y1 - 2010/9

N2 - This paper presents for the first time the application of Cartesian Genetic Programming to the evolution of machine code for a simple implementation of a MOVE processor. The effectiveness of the algorithm is demonstrated by evolving machine code for a 4-bit multiplier with three different levels of parallelism. The results show that 100% successful solutions were found by COP and by further optimising the size of the solutions, it is possible to find efficient implementations of the 4-bit multiplier that have the potential to be "human competitive". Further analysis of the results revealed that the structure of some solutions followed a known general design methodology.

AB - This paper presents for the first time the application of Cartesian Genetic Programming to the evolution of machine code for a simple implementation of a MOVE processor. The effectiveness of the algorithm is demonstrated by evolving machine code for a 4-bit multiplier with three different levels of parallelism. The results show that 100% successful solutions were found by COP and by further optimising the size of the solutions, it is possible to find efficient implementations of the 4-bit multiplier that have the potential to be "human competitive". Further analysis of the results revealed that the structure of some solutions followed a known general design methodology.

UR - http://www.scopus.com/inward/record.url?scp=78049375361&partnerID=8YFLogxK

M3 - Article

VL - 6274

SP - 238

EP - 249

JO - 9th International Conference on Evolvable Systems

JF - 9th International Conference on Evolvable Systems

ER -