{"title":"用于移动视频应用的0.3mW 1.4mm2运动估计处理器","authors":"S. Hiratsuka, S. Goto, T. Ikenaga","doi":"10.1109/ASSCC.2006.357862","DOIUrl":null,"url":null,"abstract":"Motion estimation (ME) is a key processing in video encoding systems. Since it requires huge computational complexity, many algorithms and LSI architectures have been proposed to reduce it. Conventional LSIs, however, are not sufficient for mobile applications which require both flexibility and low power dissipation. This paper describes an application specific instruction processor (ASIP) LSI for ME processing. It has a dedicated unit for SAD (sum of absolute difference) operations. By applying our proposed ultra-low ME algorithm named ULCMEA, it can reduce power while keeping high flexibility. A chip capable of operating at 80 MHz was fabricated using TSMC 0.18-mum CMOS technology. 15 K logic gates and 32 Kbit SRAM have been integrated into 1.4 mm chip. Typical power dissipation is 0.3-mW for QCIF 15 frame/ sec ME processing.","PeriodicalId":142478,"journal":{"name":"2006 IEEE Asian Solid-State Circuits Conference","volume":"1 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2006-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"2","resultStr":"{\"title\":\"A 0.3mW 1.4mm2 Motion Estimation Processor for Mobile Video Application\",\"authors\":\"S. Hiratsuka, S. Goto, T. Ikenaga\",\"doi\":\"10.1109/ASSCC.2006.357862\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"Motion estimation (ME) is a key processing in video encoding systems. Since it requires huge computational complexity, many algorithms and LSI architectures have been proposed to reduce it. Conventional LSIs, however, are not sufficient for mobile applications which require both flexibility and low power dissipation. This paper describes an application specific instruction processor (ASIP) LSI for ME processing. It has a dedicated unit for SAD (sum of absolute difference) operations. By applying our proposed ultra-low ME algorithm named ULCMEA, it can reduce power while keeping high flexibility. A chip capable of operating at 80 MHz was fabricated using TSMC 0.18-mum CMOS technology. 15 K logic gates and 32 Kbit SRAM have been integrated into 1.4 mm chip. Typical power dissipation is 0.3-mW for QCIF 15 frame/ sec ME processing.\",\"PeriodicalId\":142478,\"journal\":{\"name\":\"2006 IEEE Asian Solid-State Circuits Conference\",\"volume\":\"1 1\",\"pages\":\"0\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2006-12-01\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"2\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"2006 IEEE Asian Solid-State Circuits Conference\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.1109/ASSCC.2006.357862\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"2006 IEEE Asian Solid-State Circuits Conference","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/ASSCC.2006.357862","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 2
摘要
运动估计是视频编码系统中的一个关键处理环节。由于它需要巨大的计算复杂度,已经提出了许多算法和LSI架构来降低它。然而,传统的lsi对于需要灵活性和低功耗的移动应用来说是不够的。本文介绍了一种用于ME处理的专用指令处理器(ASIP) LSI。它有一个用于绝对差和操作的专用单元。采用我们提出的超低ME算法ULCMEA,可以在降低功耗的同时保持较高的灵活性。采用台积电0.18 μ m CMOS技术制备了工作频率为80 MHz的芯片。1.4 mm芯片集成了15k逻辑门和32kbit SRAM。对于QCIF 15帧/秒的ME处理,典型功耗为0.3 mw。
A 0.3mW 1.4mm2 Motion Estimation Processor for Mobile Video Application
Motion estimation (ME) is a key processing in video encoding systems. Since it requires huge computational complexity, many algorithms and LSI architectures have been proposed to reduce it. Conventional LSIs, however, are not sufficient for mobile applications which require both flexibility and low power dissipation. This paper describes an application specific instruction processor (ASIP) LSI for ME processing. It has a dedicated unit for SAD (sum of absolute difference) operations. By applying our proposed ultra-low ME algorithm named ULCMEA, it can reduce power while keeping high flexibility. A chip capable of operating at 80 MHz was fabricated using TSMC 0.18-mum CMOS technology. 15 K logic gates and 32 Kbit SRAM have been integrated into 1.4 mm chip. Typical power dissipation is 0.3-mW for QCIF 15 frame/ sec ME processing.