A Graphical Tool for Automatic Parallelization and Scheduling of Programs on Multiprocessors

In this paper, we introduce an experimental software tool called CASCH (Computer Aided SCHeduling) for automatic parallelization and scheduling of applications to parallel processors. CASCH transforms a sequential program to a parallel program through automatic task graph generation, scheduling, mapping, communication, and synchronization primitives insertion. The major strength of CASCH is its extensive library of state-of-the-art scheduling and mapping algorithms reported in the recent literature. Using these algorithms, a practitioner can choose the most suitable one for generating the shortest schedule for the application at hand. Furthermore, the scheduling algorithms can be interactively analyzed, tested and compared using real data on a common platform with various performance objectives. CASCH with its graphical interface is useful for both novice and expert programmers of parallel machines, and can serve as a teaching and learning aid for understanding scheduling and mapping algorithms.