{"title":"人工神经网络节能内存和传感器计算的混合信号电路和体系结构","authors":"Bongjin Kim","doi":"10.1109/SOCC46988.2019.1570571922","DOIUrl":null,"url":null,"abstract":"Von Neumann architecture is recently facing a critical challenge with the high demands of energy-efficient computing hardware for a variety of machine learning tasks such as image classification. In particular, battery-operated mobile devices with limited power budget cannot process artificial neural networks (ANNs) with relatively low complexity by using traditional digital circuits and architectures. The key challenge with the traditional Von Neumann architecture is its energy-inefficient data access between memory and processor. Recently, in-memory computing architecture has gained significant attention as an alternative, especially for running mobile artificial intelligence applications. The memory access energy has been drastically reduced by using local in-memory processing elements which directly use the data stored in local memory. To further improve the efficiency, the processor based on mixed-signal circuits instead of conventional digital circuits have recently been actively researched. However, mixed-signal circuits have several critical drawbacks, including nonlinearity, PVT variation, and the overhead of ADC/DAC for interfacing with the external digital domain. In this work, we first review the recent mixed-signal circuits and architectures for in-memory computing using different embedded memories. In addition, we introduce the concept of in-sensor computation for integrating partial computing units in an image-sensor array using low-power mixed-signal circuit techniques.","PeriodicalId":253998,"journal":{"name":"2019 32nd IEEE International System-on-Chip Conference (SOCC)","volume":"92 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2019-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"2","resultStr":"{\"title\":\"Mixed-Signal Circuits and Architectures for Energy-Efficient In-Memory and In-Sensor Computation of Artificial Neural Networks\",\"authors\":\"Bongjin Kim\",\"doi\":\"10.1109/SOCC46988.2019.1570571922\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"Von Neumann architecture is recently facing a critical challenge with the high demands of energy-efficient computing hardware for a variety of machine learning tasks such as image classification. In particular, battery-operated mobile devices with limited power budget cannot process artificial neural networks (ANNs) with relatively low complexity by using traditional digital circuits and architectures. The key challenge with the traditional Von Neumann architecture is its energy-inefficient data access between memory and processor. Recently, in-memory computing architecture has gained significant attention as an alternative, especially for running mobile artificial intelligence applications. The memory access energy has been drastically reduced by using local in-memory processing elements which directly use the data stored in local memory. To further improve the efficiency, the processor based on mixed-signal circuits instead of conventional digital circuits have recently been actively researched. However, mixed-signal circuits have several critical drawbacks, including nonlinearity, PVT variation, and the overhead of ADC/DAC for interfacing with the external digital domain. In this work, we first review the recent mixed-signal circuits and architectures for in-memory computing using different embedded memories. In addition, we introduce the concept of in-sensor computation for integrating partial computing units in an image-sensor array using low-power mixed-signal circuit techniques.\",\"PeriodicalId\":253998,\"journal\":{\"name\":\"2019 32nd IEEE International System-on-Chip Conference (SOCC)\",\"volume\":\"92 1\",\"pages\":\"0\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2019-09-01\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"2\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"2019 32nd IEEE International System-on-Chip Conference (SOCC)\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.1109/SOCC46988.2019.1570571922\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"2019 32nd IEEE International System-on-Chip Conference (SOCC)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/SOCC46988.2019.1570571922","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
Von Neumann架构最近面临着一个严峻的挑战,即对各种机器学习任务(如图像分类)的高能效计算硬件的高要求。特别是,电池供电的移动设备在有限的功率预算下,无法使用传统的数字电路和架构处理复杂度相对较低的人工神经网络(ann)。传统冯·诺依曼架构的主要挑战是其在存储器和处理器之间的低能耗数据访问。最近,内存计算架构作为一种替代方案获得了极大的关注,特别是在运行移动人工智能应用程序时。通过使用直接使用存储在本地存储器中的数据的本地内存处理元件,大大减少了存储器访问能量。为了进一步提高效率,近年来人们积极研究以混合信号电路代替传统数字电路的处理器。然而,混合信号电路有几个关键的缺点,包括非线性、PVT变化和ADC/DAC与外部数字域接口的开销。在这项工作中,我们首先回顾了最近使用不同嵌入式存储器进行内存计算的混合信号电路和架构。此外,我们引入传感器内计算的概念,利用低功耗混合信号电路技术集成图像传感器阵列中的部分计算单元。
Mixed-Signal Circuits and Architectures for Energy-Efficient In-Memory and In-Sensor Computation of Artificial Neural Networks
Von Neumann architecture is recently facing a critical challenge with the high demands of energy-efficient computing hardware for a variety of machine learning tasks such as image classification. In particular, battery-operated mobile devices with limited power budget cannot process artificial neural networks (ANNs) with relatively low complexity by using traditional digital circuits and architectures. The key challenge with the traditional Von Neumann architecture is its energy-inefficient data access between memory and processor. Recently, in-memory computing architecture has gained significant attention as an alternative, especially for running mobile artificial intelligence applications. The memory access energy has been drastically reduced by using local in-memory processing elements which directly use the data stored in local memory. To further improve the efficiency, the processor based on mixed-signal circuits instead of conventional digital circuits have recently been actively researched. However, mixed-signal circuits have several critical drawbacks, including nonlinearity, PVT variation, and the overhead of ADC/DAC for interfacing with the external digital domain. In this work, we first review the recent mixed-signal circuits and architectures for in-memory computing using different embedded memories. In addition, we introduce the concept of in-sensor computation for integrating partial computing units in an image-sensor array using low-power mixed-signal circuit techniques.