A Type System for Parallel Components
Source: Universidade Federal do Ceara
Multi-core processors have already made parallel computing a mainstream technology, but High Performance Computing (HPC) applications that run on clusters and grids have already attracted the investments of the software industry. The key for reaching peak performance is the knowledge of how to apply HPC techniques for parallel programming by looking at the particular features of the parallel computing architecture. With the raising of complexity and scale of HPC applications [Post and Votta 2005], HPC developers now demands for software engineering artifacts to develop HPC software [Sarkar et al. 2004]. Unfortunately, parallel programming is still hard to be incorporated into usual software development platforms.