Cache Reconfiguration Using Machine Learning for Vulnerability-aware Energy Optimization

Abstract

Dynamic cache reconfiguration has been widely explored for energy optimization and performance improvement for single-core systems. Cache partitioning techniques are introduced for the shared cache in multicore systems to alleviate inter-core interference. While these techniques focus only on performance and energy, they ignore vulnerability due to soft errors. In this article, we present a static profiling based algorithm to enable vulnerability-aware energy-optimization for real-time multicore systems. Our approach can efficiently search the space of cache configurations and partitioning schemes for energy optimization while task deadlines and vulnerability constraints are satisfied. A machine learning technique has been employed to minimize the static profiling time without sacrificing the accuracy of results. Our experimental results demonstrate that our approach can achieve 19.2% average energy savings compared with the base configuration, while drastically reducing the vulnerability (49.3% on average) compared to state-of-the-art techniques. Furthermore, the machine learning technique enabled more than 10x speedup in static profiling time with a negligible prediction error of 3%.

Dynamic machine learning‐based heuristic energy optimization approach on multicore architecture

“…The adjacent set of tasks is clearly determined for each combination to minimize the results in the problem. Therefore, the conditions of the dynamic program can be represented in Equation (11) as…”

“…where k i represents max(k crt , Z i ) and this set of equations is built to form all the subproblems with the consecutive set of tasks. The function of recursive elements is determined for Equation (11) and it can be defined in the form of Equation ( 12) as…”

“…As shown in Figure 1, it is one of the very equivalent applications, which will help to evaluate and monitor the samples in the data to identify the health condition if there is any emergency period. 11 Concentrating toward the stress level of these sensor devices on multicore architecture will completely help manage the stress points and also it will help improve the quality of the patient life. Since the data will be monitoring day to day, this wearable devices will completely help if there is any situation during emergency on multicore chips.…”