Scheduling Algorithms for Effective Thread Pairing on Hybrid Multiprocessors

With the latest high-end computing nodes combining shared-memory multiprocessing with hardware multithreading, new scheduling policies are necessary for workloads consisting of multithreaded applications. The use of hybrid multiprocessors presents schedulers with the problem of job pairing, i.e. deciding which specific jobs can share each processor with minimum performance penalty, by running on different execution contexts. Therefore, scheduling policies are expected to decide not only which job mix will execute simultaneously across the processors, but also which jobs can be combined within each processor. This paper addresses the problem by introducing new scheduling policies that use run-time performance information to identify the best mix of threads to run across processors and within each processor. Scheduling of threads across processors is driven by the memory bandwidth utilization of the threads, whereas scheduling of threads within processors is driven by one of three metr...

• Type: Conference
• Year: 2005
• Where: IPPS
• Authors: Robert L. McGregor, Christos D. Antonopoulos, Dimitrios S. Nikolopoulos
• Keywords: Bus Transaction Rate, Distributed And Parallel Computing, IPPS 2005, Scheduling Policies, Thread Pairing