Flexible Memory Allocation in Kinetic Monte Carlo Simulations

We introduce two new algorithms for Kinetic Monte Carlo simulations: the minimal and flexible allocation algorithms. The theory and computational challenges associated with K.M.C. simulations are briefly discussed. We outline the simple cubic, solid-on-solid model of epitaxial growth and analyze four methods for its simulation: the linear search, standard inverted list, minimal allocation, and flexible allocation algorithms. We then implement these algorithms, analyze their performances, and discuss implications of the results.