{"title":"Low energy optimization for MPEG-4 video encoder on ARM-based mobile phones","authors":"P. Pakdeepaiboonpol, S. Kittitornkun","doi":"10.1109/ISWPC.2006.1613618","DOIUrl":null,"url":null,"abstract":"Most compiler optimization techniques concern most about speed. In this paper, we present two high-level power/energy optimization methods for ARM-based battery-powered embedded multimedia systems, e.g. mobile phones, pocket PCs, personal multimedia systems, etc. The experiments using MPEG-4 simple profile level 0 (SP@L0) video encoder on ARM920T with two QCIF video sequences 15 fps, 24 kbps show that the proposed techniques can complement the existing speed-oriented ones to achieve lower energy/power consumption up to 5.7% relative to all ARM C++ optimization levels despite the 16-KB instruction and 16-KB data caches of ARM 920T core.","PeriodicalId":145728,"journal":{"name":"2006 1st International Symposium on Wireless Pervasive Computing","volume":"30 Suppl 1 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2006-04-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"3","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2006 1st International Symposium on Wireless Pervasive Computing","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/ISWPC.2006.1613618","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 3
Abstract
Most compiler optimization techniques concern most about speed. In this paper, we present two high-level power/energy optimization methods for ARM-based battery-powered embedded multimedia systems, e.g. mobile phones, pocket PCs, personal multimedia systems, etc. The experiments using MPEG-4 simple profile level 0 (SP@L0) video encoder on ARM920T with two QCIF video sequences 15 fps, 24 kbps show that the proposed techniques can complement the existing speed-oriented ones to achieve lower energy/power consumption up to 5.7% relative to all ARM C++ optimization levels despite the 16-KB instruction and 16-KB data caches of ARM 920T core.