As concern grows for the development of clean and sustainable energy sources and the advances in computation, engineering, and materials science converge, emphasis is placed in the development of more efficient materials.  A problem in material science is that nucleation is poorly understood.  Nucleation is the process by which the first stable embryo of a precipitate forms.  Precipitates grow and then coarsen to form the microstructure of the material, which largely influences material properties.   Simulations exist for the later stages of growth and coarsening but not for the nucleation step.  Simulating nucleation is difficult as a successful nucleation event may involve millions of atom hops and exceed the limits of standard kinetic Monte Carlo simulations.  We hypothesize that a timescale separation between the atom hopping time and the nucleation time can be exploited to understand nucleation mechanisms in metastable solids at realistic supersaturations.