On-line HW/SW partitioning and co-scheduling in reconfigurable computing systems

Abstract

Reconfigurable computing systems allow executing tasks in a true multitasking manner. Such systems share the reconfigurable device and processing unit as computing resources which leads to highly dynamic allocation situations. To manage such systems at runtime, a reconfigurable operating system is needed. The main part of this operating system is resource management unit which performs HW/SW partitioning, co-scheduling and placement of hardware tasks at run-time. In this paper, we present a heuristic for on-line integrated HW/SW partitioning and co-scheduling. We focus on on-line, non real-time and non-preemptive systems. The main characteristic of our heuristic is strong nexus between partitioning, scheduling and placement. Our heuristic prioritizes the arrived tasks according to different important parameters and partitions the sorted tasks according to their earliest finish time (EFT) on software and hardware processing units. A large variety of experiments have been conducted on the proposed algorithm using synthetic tasks. Obtained results show considerable benefits of this algorithm.