A universal approach for task scheduling for distributed memory multiprocessors

We present a static one-step list scheduling technique for scheduling a task graph onto a distributed memory multiprocessor taking into account the interconnection constraints and channel conflicts. We use a priority list to directly schedule nodes in a task graph onto the processor architecture. The scheme departs from conventional schedulers in its use of a "bottom-up" approach for scheduling the task graph nodes. This scheduling technique is applicable to any type of processor architecture and routing strategy. Experimental results indicate the performance advantages of the scheduler.<
>