Energy-Efficient Execution of Streaming Task Graphs with Parallelizable Tasks on Multicore Platforms with Core Failures

Real-time applications often take the form of streaming applications, where a stream of inputs such as camera images is processed by an application represented as a task graph. The workload together with the required throughput often necessitates processing on a multicore system and also demands parallelization of large tasks. We extend a scheduling algorithm for such applications, originally devised to handle varying task workloads, to also cover varying core count, e.g. caused by failure of a core. We use frequency scaling to accelerate processing when the necessity to re-execute tasks from the crashed core arises, to maintain throughput. We evaluate the algorithm by scheduling synthetic task graphs that represent corner cases and a real streaming application.