Parallel programming is a hot topic among scientists and engineers. The number of parallel machines available to them is constantly increasing. These powerful machines provide a new research platform for engineers and scientists. In order to exploit all their power, new programming techniques are needed. Fortunately, several parallel programming tools have appeared to make this complex programming endeavor much easier. We analyze paths to parallel programming from traditional scientific programming through parallelization of a molecular simulation application. We hope this paper encourages scientists and engineers at Cornell's Theory Center to take the step towards parallelization.