13th European Conference on Genetic Programming, İstanbul, Turkey, 7 - 09 April 2010, vol.6021, pp.325-326
In previous work, we have demonstrated that it is possible to use Genetic Programming to minimise the resource consumption of software, such as its power consumption or execution time. In this paper, we investigate the extent to which Genetic Programming can be used to gain fine-grained control over software timing. We introduce the ideas behind our work, and carry out experimentation to find that Genetic Programming is indeed able to produce software with unusual and desirable timing properties, where it is not obvious how a manual approach could replicate such results. In general, we discover that Genetic Programming is most effective in controlling statistical properties of software rather than precise control over its timing for individual inputs. This control may find useful application in cryptography and embedded systems.