In this chapter we will focus on molecular dynamics (MD) simulations with large numbers of atoms ( N > 1000). Typically, ab initio or F irst P rinciples M olecular D ynamics (FPMD) calculations are performed with smaller clusters of say 100 or fewer atoms. This choice is largely due to limitations in computing resources. The simulation procedure we will explore in this chapter is known as E mpirical P air- P otential M olecular D ynamics (EPPMD). EPPMD does not rely upon estimating the forces between atoms with D ensity F unctional T heory (DFT) as an approximation to the solution of the quantum mechanical problem, which is the computationally costly aspect of FPMD. Rather, EPPMD estimates forces between atoms from an empirical parameterization of a classical description of the potential energy, whose spatial derivative captures the pairwise-additive attractive and repulsive forces between all atoms. Because the interatomic force calculation is classical, its computational cost is minimal, and this encourages the application of EPPMD to simulations involving large numbers of atoms. Additionally, whereas FPMD simulations are typically run for durations on the order of a few picoseconds, EPPMD simulations may be extended to much longer durations (2–10 nanoseconds), which permits investigation of transport properties such as shear viscosity and phonon thermal conductivity (Fig. 1⇓) using Green-Kubo theory (Kubo 1966). The longer durations of EPPMD runs (with up to order 107 time steps) are again a consequence of the computational efficiency derived from the classical force field approximation. In this chapter we are going to refer to long duration molecular dynamics simulations involving large numbers of atoms as L arge S cale S imulations (LSS). Figure 1. Typical time scales for various molecular dynamics calculations. The value labeled as “Debye frequency” refers to the reciprocal of (3 N /4π V ) …