A direct coupled cluster algorithm for massively parallel computers

A closed-shell coupled cluster program specially designed to run efficiently on massively parallel computers is presented. The input/output bottleneck present in an earlier implementation has been circumvented by solving the coupled cluster equations in a direct manner. Sample calculations on glycine, cytosine and serine have been run on the CRAY T3D and T3E. The results show good scalability up to 256 processors.