David Schwenke

• Research Scientist, NASA Ames Research Center (ARC)
• Computational Physics Branch
• Please enable Javascript to see email address.

Scientific computing on High End Computing Platforms. Carried pioneering work in Computational Chemistry at the University of Minnesota where the first codes were vectorized for the CDC Cyber205 in 1984. With the next generation machine, the Cray-1, asynchronous I/O enabled efficient massive simulations that were state of the art at the time. Similar pioneering work was carried on the Cray-2. I Joined ARC in 1987 and have been here ever since. With the emergence of parallel shared memory systems in 1990’s, started coding using OpenMP for parallelization. Joined the NAS division in late 1990’s. During the acquisition of Columbia, planning for a parallel electronic structure code, MPEC, was started. This pioneering code used shmem for parallelization and was built from the ground up with large numbers of processors in mind. Ultimately, MPEC was capable of good scaling and made possible calculations that were previously unreachable with single processor technology. My current efforts are focused on optimizing my Schrodinger equation solver for the Pleaides architecture. I have now been working on MPEC version 2 that uses MPI for parallelization. In spite of the set back going from shmem to MPI, MPEC version 2 has turned into a much better code.