An adaptive scheduling scheme for parallel programs on shared-memory multicores

Extracting performance from modern multicore architectures requires that parallel sections be divided into many threads of execution. In order to fully utilize these threads effectively, load balancing has become one of the most important factors that affect the performance of applications on multicores. In this paper, by evaluating the parallel programs, we show that the threads that belong to the same application can exhibit a poorly balancing performance. In order to extract the full efficiency available from the imbalance of these threads, we propose an adaptive scheduling scheme which can adjust the workload chunk size for scheduling among the threads during the applications running. Our experimental evaluation indicates that the proposed adaptive scheduling scheme yields 8% performance improvement over a static scheduling scheme and 4% over a dynamic scheme on average, respectively.