Choonghoon Park, Jong-Kab Park, Youngtaek Lee, B. Park, Jungwook Kim, Donghee Han, Chulmin Jo, Woonhaing Hur
{"title":"A new DVFS algorithm to minimize energy consumption on system-on-chip architecture and electrical characteristics","authors":"Choonghoon Park, Jong-Kab Park, Youngtaek Lee, B. Park, Jungwook Kim, Donghee Han, Chulmin Jo, Woonhaing Hur","doi":"10.1109/ICCE53296.2022.9730423","DOIUrl":null,"url":null,"abstract":"Over the decades, lots of power management methods have been devised to increase battery life time in mobile devices. Among them, DVFS (Dynamic Voltage and Frequency Scaling) algorithms have been introduced to minimize energy consumption while meeting the required performance at the same time. Generally, most DVFS algorithms aim to minimize energy consumption without hurting QoS (Quality of Service) by selecting the lowest frequency which satisfies the performance required by workloads. However, it becomes more difficult to achieve energy efficiency with the existing DVFS algorithms on highly complex systems, such as de-coupling between operating frequency and voltage, and increasing static power due to the fine process trend. Therefore, the need for an entirely new DVFS algorithm has come to the fore. The new algorithm is designed to be well-deployed onto such complex systems, and follow the fundamental goal - minimizing energy, not simply reducing the power. In this paper, we are going to introduce an algorithm to satisfy the requirements.","PeriodicalId":350644,"journal":{"name":"2022 IEEE International Conference on Consumer Electronics (ICCE)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2022-01-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2022 IEEE International Conference on Consumer Electronics (ICCE)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/ICCE53296.2022.9730423","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0
Abstract
Over the decades, lots of power management methods have been devised to increase battery life time in mobile devices. Among them, DVFS (Dynamic Voltage and Frequency Scaling) algorithms have been introduced to minimize energy consumption while meeting the required performance at the same time. Generally, most DVFS algorithms aim to minimize energy consumption without hurting QoS (Quality of Service) by selecting the lowest frequency which satisfies the performance required by workloads. However, it becomes more difficult to achieve energy efficiency with the existing DVFS algorithms on highly complex systems, such as de-coupling between operating frequency and voltage, and increasing static power due to the fine process trend. Therefore, the need for an entirely new DVFS algorithm has come to the fore. The new algorithm is designed to be well-deployed onto such complex systems, and follow the fundamental goal - minimizing energy, not simply reducing the power. In this paper, we are going to introduce an algorithm to satisfy the requirements.
几十年来,人们设计了许多电源管理方法来延长移动设备的电池寿命。其中引入了DVFS (Dynamic Voltage and Frequency Scaling,动态电压频率缩放)算法,在满足性能要求的同时最大限度地降低能耗。通常,大多数DVFS算法的目标是通过选择满足工作负载所需性能的最低频率,在不损害QoS(服务质量)的情况下最小化能耗。然而,现有的DVFS算法在高度复杂的系统上难以实现能效,如工作频率与电压的解耦,以及由于精细化趋势导致的静态功率的增加。因此,对全新的DVFS算法的需求已经浮出水面。新算法被设计成可以很好地部署到这种复杂的系统中,并遵循基本目标-最小化能量,而不仅仅是降低功率。在本文中,我们将介绍一种算法来满足这些要求。