On the Scalability of Real-Time Scheduling Algorithms on Multicore Platforms: A Case Study

Multicore platforms are predicted to become significantlylarger in the coming years. Given that real-time workloadswill inevitably be deployed on such platforms, the scalabilityof the scheduling algorithms used to support such workloadswarrants investigation. In this paper, this issue is consideredand an empirical evaluation of several global andpartitioned scheduling algorithms is presented. This evaluationwas conducted using a Sun Niagara multicore platformwith 32 logical CPUs (eight cores, four hardware threadsper core). In this study, each tested algorithm proved to bea viable choice for some subset of the workload categoriesconsidered.