PARALLEL PSEUDOSPECTRAL ELECTRONIC-STRUCTURE - I - HARTREE-FOCK CALCULATIONS

We present an outline of the parallel implementation of our pseudospec tral electronic structure program, Jaguar, including the algorithm and timings for the Hartree-Fock and analytic gradient portions of the pr ogram. We also present the parallel algorithm and timings for our Lanc zos eigenvector refinement code and demonstrate that its performance i s superior to the ScaLAPACK diagonalization routines. The overall effi ciency of our code increases as the size of the calculation is increas ed, demonstrating actual as well as theoretical scalability. For our l argest test system, alanine pentapeptide [818 basis functions in the c c-pVTZ(-f) basis set], our Fock matrix assembly procedure has an effic iency of nearly 90% on a 16-processor SP2 partition. The SCF portion f or this case (including eigenvector refinement) has an overall efficie ncy of 87% on a partition of 8 processors and 74% on a partition of 16 processors. Finally, our parallel gradient calculations have a parall el efficiency of 84% on 8 processors for porphine (430 basis functions ). (C) 1998 John Wiley & Sons, Inc.