Bubble task: A dynamic execution throttling method for multi-core resource management

Abstract

Memory bandwidth is a major resource which is shared among all CPU cores. The development speed of memory bandwidth cannot catch up with the increasing number of CPU cores. Thus, the contention for occupying more memory bandwidth among concurrently executing tasks occurs. In this paper, we have presented Bubble Task method which mitigates memory contention via throttling technique. We made a memory contention modeling for dynamically deciding throttling ratio and implemented both software and hardware versions to present trade-off between fine-grained adjustment and stable fairness. Bubble Task can lead to performance improvement in STREAM benchmark suite which is one of the most memory hungry benchmark by 21% and fairness in memory bandwidth sharing among SPEC CPU 2006 applications which have different memory access patterns. © Springer International Publishing Switzerland 2015.