| Summary: | 碩士 === 國立臺灣科技大學 === 電機工程系 === 106 === Modern mobile devices are equipped with heterogeneous multicore processors which integrate asymmetric CPU cores and GPUs. More cores require additional power consumption and produce more heat, which can result in performance degradation due to thermal throttling. To address this issue, this paper proposes a QT-adaption engine to monitor current temperature and QoS, and derive a performance and thermal model (QT-model) through a run-time learning mechanism (QT-learning) to balance dynamic workloads and dynamic thermal behavior. Based on the derived QT-model, the QT-adaption engine migrates threads among cores using the proposed CT-aware scheduler to ensure high QoS, and uses a self-adaption governor to meet the temperature constraint for system robustness. The concept is implemented on a commercial LG Nexus 5x and evaluated using real world applications. Results show the proposed approach can improve QoS by up to 35% FPS compared to other current methods while meeting temperature constraints.
|
|---|
