SEMPA: software engineering for parallel scientific computing

The Sempa project brings together researchers from computer science, mechanical engineering, and numerical analysis to define software-engineering methods for the parallelization of existing large-scale software packages in scientific computing. The parallel implementation of TfC, an industrial state-of-the-art computational-fluid-dynamics simulation program, serves as the central case study for defining and evaluating these methods. Sempa researchers have successfully implemented and tested a portable parallel version of TfC based on message-passing standards (PVM and MPI). In addition, Sempa researchers have demonstrated the potential of new languages and programming paradigms such as data parallelism and object orientation by reimplementing the algebraic multigrid solver-a key module in TfC-in Fortran 90, HPF, and C++. Networks of workstations have become very attractive as low-cost platforms. Efficiently using their resources for production runs of parallel software requires automated resource management that optimizes resource utilization for parallel batch jobs without impeding interactive use. Therefore, the Sempa project has also implemented a resource manager for batch execution of PVM programs