PROJECT TITLE :

Predicting Cross-Core Performance Interference on Multicore Processors with Regression Analysis

ABSTRACT:

Despite their widespread adoption in cloud computing, multicore processors are heavily underneath-utilized in terms of computing resources. To avoid the potential for negative and unpredictable interference, co-location of a latency-sensitive application with others on the identical multicore processor is disallowed, leaving several cores idle and inflicting low machine utilization. To enable co-location while providing QoS guarantees, it's challenging but important to predict performance interference between co-located applications. We tend to observed that the performance degradation of an application can be represented as a piecewise predictor function of the aggregate pressures on shared resources from all cores. Based mostly on this observation, we have a tendency to propose to adopt regression analysis to create a predictor perform for an application. Furthermore, the prediction model thus obtained for an application is able to characterize its contentiousness and sensitivity. Validation using a giant range of single-threaded and multi-threaded benchmarks and 9 real-world datacenter applications on two completely different platforms shows that our approach is additionally precise, with a mean error not exceeding 0.4 p.c.