The goal of our research is touse computational methods to predict the structures and free energies of macromolecules and their complexes. To achieve this goal requires not only clever algorithms and methods to search both intramolecular and intermolecular conformational space, but methods to accurately evaluate the free energies of each of the structures generated. Our groups have been at the forefront of both searching and free energy evaluation and have made much progress on both fronts. However, our research progress is severely hampered by a lack of adequate computational resources and the proposed multiprocessor computer system would greatly enhance the research productivity of our research groups.