Performance Evaluation of Reusing Based Scheduling in On-line Reconfigurable Computing Systems
Abstract
Reconfiguration overhead is an important obstacle that limits the performance of on-line scheduling algorithms in reconfigurable computing systems and increases the overall execution time. Configuration reusing (task reusing) can decrease reconfiguration overhead considerably, particularly in periodic applications or the applications in which the probability of tasks recurrence is high. In this paper, we improve reusing technique in order to reduce reconfiguration overhead and decrease total execution time of the tasks. In reusing method, input tasks are divided into significant and non-significant tasks. Significant tasks are preserved on the RPU in order to be reused in near future. The main contributions in this paper include new task significance calculation and performance evaluation of resizing the RPU partitions. A large variety of experiments has been conducted on the proposed algorithm. Obtained results show when the recurrence of tasks is 40%, we have up to 14% improvement in overall execution time of resizable reusing based scheduling comparing reusing without resizing, and up to 25% improvement comparing scheduling without reusing.
Keywords
Reconfigurable Computing, On-Line Scheduling, Configuration Reusing, RPU Partitioning, Replacement Management, Probability of Recurrence
