Machine learning for power, energy, and thermal management on multi-core processors: A survey.

Abstract

Due to the high integration density and roadblock of voltage scaling, modern multi-core processors experience higher power densities than previous technology scaling nodes. When unattended, this issue might lead to temperature hot spots, that in turn may cause non-uniform aging, accelerate chip failure, impair reliability, and reduce the performance of the system. This paper presents an overview of several research efforts that propose to use machine learning techniques for power and thermal management on single-core and multi-core processors. Traditional power and thermal management techniques rely on a certain a-priori knowledge of the chip's thermal model, as well as information of the workloads/applications to be executed (e.g., transient and average power consumption). Nevertheless, these a-priori information is not always available, and even if it is, it cannot reflect the spatial and temporal uncertainties and variations that come from the environment, the hardware, or from the workloads/applications. Contrarily, techniques based on machine learning can potentially adapt to varying system conditions and workloads, learning from past events in order to improve themselves as the environment changes, resulting in improved management decisions.