A Survey of Accelerator Architectures for Deep Neural Networks

Recently, due to the availability of big data and the rapid growth of computing power, artificial intelligence (AI) has regained tremendous attention and investment. Machine learning (ML) approaches have been successfully applied to solve many problems in academia and in industry. Although the explosion of big data applications is driving the development of ML, it also imposes severe challenges of data processing speed and scalability on conventional computer systems. Computing platforms that are dedicatedly designed for AI applications have been considered, ranging from a complement to von Neumann platforms to a “must-have” and stand-alone technical solution. These platforms, which belong to a larger category named “domain-specific computing,” focus on specific customization for AI. In this article, we focus on summarizing the recent advances in accelerator designs for deep neural networks (DNNs)—that is, DNN accelerators. We discuss various architectures that support DNN executions in terms of computing units, dataflow optimization, targeted network topologies, architectures on emerging technologies, and accelerators for emerging applications. We also provide our visions on the future trend of AI chip designs.     

文化大数据与智能设计平台综述

目的 基于对文化大数据平台和智能设计平台的发展现状研究,阐述多源异构融合文化大数据平台建设面临的挑战,提出文化大数据智能设计平台的系统架构,在传承和保护文化的同时,探索文化数据在设计领域的再开发与再利用.方法 通过对相关文献的收集、整理和分析,了解文化大数据平台和智能设计平台的国内外研究进展,综合对比分析其技术原理、研究难点与挑战,并提出文化大数据智能设计平台的基础架构.结论 文化大数据平台的发展促进了文化的保护利用,但是仍存在利用率低、使用较困难等问题;人工智能技术在设计领域的应用,为设计学科引入了新的设计方法和路径,可以有效提升设计效率与质量,但当前的智能设计工具及平台仍然存在功能较为单一、缺乏灵活性等问题,有待进一步提升;文化大数据智能设计平台的研发,可以提升文化大数据的利用效率,探索文化大数据在设计不同阶段、不同方向的应用,提升设计作品的文化底蕴与质量.     

Analysis of Wireless Backhaul Networks Based on Aerial Platform Technology for 6G Systems

As next generation communication technologies emerge, new high data rate applications and high-definition large-screen video streaming have become very popular. As a result, network traffic has been increasing so much that existing backhaul networks soon will not be able to support all traffic demands. To support these needs in future 6G mobile systems, the establishment of an additional backhaul wireless network is considered essential. As one of the solutions, a wireless backhaul network based on an aerial platform has been proposed. In order to explore the potential of aerial platforms as wireless backhaul networks, in this paper, the categories for wireless backhaul networks based on aerial platforms are investigated. This paper includes a survey of the definitions and characteristics of low altitude platforms (LAPs) and high altitude platforms (HAPs), as well as channel models according to the atmosphere. For wireless backhaul network designs based on aerial platforms, altitude and platform selection options, deployment options, energy issues, and security based on target location and performance were considered in the analysis and investigation.     

Comparisons of technological innovation capabilities in the solar photovoltaic industries of Taiwan, China, and Korea

This paper investigates the technological innovation capabilities of the three Asian latecomers—namely Taiwan, China, and Korea—in the emergent solar photovoltaic industry. For this study, I deploy a new dataset of 75,540 solar photovoltaic patents taken out by Taiwan, China and Korea over the period of 31 years (1978–2008) and analyse the evolving technological innovation capabilities revealed in these patents using a set of four technology platforms that I constructed. This study demonstrates the patent portfolios of the three latecomers and explores what extent have the Taiwanese, Chinese, and Korean followers developed their technological innovation capabilities so as to surpass the US, German, and Japan and acquire the leading production positions; and how the variations of technological innovation capabilities among the major producers influence their business activities in the global solar photovoltaic industry. The results show the various strategies adopted by Taiwan, China, and Korea to develop their solar photovoltaic industries, reflect their different national innovation systems involved, and response to the current trends of technology development in the global solar photovoltaic industry.     

Dynamic patterns of technology collaboration: a case study of the Chinese automobile industry, 1985–2010

To investigate patterns of technology collaboration within the Chinese automobile industry, this study employs a unique dataset of patent applications that reveal a record of 64,938 collaborative relations in the industry during the period from 1985 to 2010. Our results indicate that over 60 % of the total collaborations were conducted after China entered the WTO. The invention and utility types of patents account for 98 % of the total collaborations throughout the sample period. Using a network analysis method, we find that the key differences between domestic enterprises collaborating with indigenous enterprises (DD collaboration) and with foreign firms (DF collaboration) are in patent types and technology domains. The DF network is also denser and more centralized than the DD network, although the amount of nodes and links of the DD network is greater than that of the DF collaboration network. The analysis and visualization of the collaboration networks and corresponding largest components reveal that a large number of domestic enterprises prefer to collaborate with top global automobile manufacturers. We also find that a number of universities have become key players in the collaborations among industry, universities and research institutes. This study provides a deeper understanding of technology collaborations from various perspectives and also highlights several avenues for future research.     

基于区块链技术的空调产品电子病历共享方案

随着空调领域大数据和智能化的飞速发展,良好的数据管理模式变得十分重要.本文首次提出一种将区块链作为底层技术的空调产品电子病历共享方案,旨在构建一个去中心化和第三方信任的空调病历信息共享平台.通过运用哈希计算、非对称加密和全网共识等技术,达到对空调故障病历数据的全网可信,突破个人、企业之间的信任壁垒.研究表明:基于区块链...     

智慧停车场在线计量云平台分析与设计研究

为提高停车场电子计时收费装置的检定效率和解决其在技术监督管理现状中存在的问题,提出运用现代云计算理念和大数据技术,建立以在线监测和校准系统为核心应用的停车场智能管理系统公共监管和服务云平台。在可行性研究的基础上,提出该系统的设计架构;在需求分析的基础上,提出智慧停车场客户端单元、在线检测单元、信息发布单元和行政管理单元等功能模块的概要设计方案;详细描述云平台的核心应用——在线计量检测的实现原理与过程。该系统用互联网技术来实现停车场计时收费系统的在线计量检定,经实践证明其改变传统停车场电子计时收费系统的计量检测模式,根据云端虚拟服务器的网络并发响应能力,检测效率理论上可以提高数千倍;实现停车信息、安全、检测和监督等综合服务集成到平台上,最大限度为管理部门和公众提供多个层面的信息服务。     

Development and Application of Big Data Platform for Garlic Industry Chain

In order to effectively solve the problems which affect the stable and healthy development of garlic industry, such as the uncertainty of the planting scale and production data, the influence factors of price fluctuation is difficult to be accurately analyzed, the difficult to predict the trend of price change, the uncertainty of the market concentration, and the difficulty of the short-term price prediction etc. the big data platform of the garlic industry chain has been developed. Combined with a variety of data acquisition technology, the information collection of influencing factors for garlic industry chain is realized. Based on the construction of the big data technology platform, the real-time synchronous acquisition, efficient storage and analysis of the planting, market, storage, processing, export and logistics information in five provinces and seven counties are realized. The application of the big data platform for garlic industry chain has realized the accurate acquisition of garlic planting area, the price and trend of market circulation and the information of export information, analyzed the fluctuation regulation of garlic price, and also realized the short-term precision prediction of garlic price.     

Technology–industry networks in technology commercialization: evidence from Korean university patents

Although there is increasing interest in policy issues on university patents, studies hitherto have focused on certain limited factors or case studies. By using a two-mode network analysis, this study identifies idiosyncratic patterns and differences in technology–industry networks between the two groups of Korean university patents—commercialized and non-commercialized. We collected patent data including bibliographic information from Korean universities that have run a patent management advisor dispatch program since 2005. Then, network analysis and analysis of variance for the two groups were conducted to investigate the group differences. We found that the structure of the technology–industry network was significantly more direct and simpler for commercialized than for non-commercialized patents. Specifically, we found that both direct and indirect linkages between technology and related industry were more complex for the non-commercialized group than for the commercialized one: the direct linkage was stronger for the commercialized than for the non-commercialized group. Our study suggests an important aspect of technology commercialization from the perspective of the inherent characteristics of patents, which is at variance with the evolutionary approaches of previous studies.     

互联网背景下制造业创新体系建设研究

以互联网和大数据为特征的新一代信息技术深刻影响着制造业创新体系建设。本文分析了互联网背景下,制造业创新体系在共性技术供给、技术供给方式和产学研创新体系建设方面的建设需求,提出了建设共性技术研发基地、提供多元化技术供给平台以促进创新网络建设和创新生态培育。     

基于B2B2C的网络订餐电子商务平台设计与实现

基于B2B2C网络订餐电子商务平台,利用计算机技术、网络技术和远程通信技术,实现订餐企业商务(买卖)过程中的电子化、数字化和网络化,为快餐企业提供了一个信息化管理解决方案.在统一平台上规范了快餐行业的管理模式标准和业务需求,完成了网上订餐电子商务平台、订餐供应商管理信息系统、平台运营管理系统的设计.主要介绍网络订餐电子商务平台的开发与设计过程,阐述了系统的功能需求、主要技术要点和设计实现过程.     

The Evolution of Network Industries: Lessons from the Conquest of the Online Frontier, 1979–95

In this paper we dissect the competitive dynamics of a network industry that has not been systematically studied in the past—the consumer‐oriented online networks of the pre‐1995 era. After briefly reconstructing the historical evolution of the industry we focus on a number of issues at the intersection of theory and data: the importance of direct and indirect network effects, tipping to a dominant player and the entry strategies adopted by two different waves of entrants. We also analyse price wars and the displacement of the leader in the first half of the 1990s. In the final section we focus on some remaining puzzles that suggest directions for further theoretical and empirical research.     

The Impact of a Radical Innovation on Business Models: Incremental Adjustments or Big Bang?

In this paper we study the impact of a radical technological innovation on business models. Do firms react by adjusting their business models incrementally, through iterative steps? Or do such innovations lead, instead, to a big bang of new innovative business models that are all adopted and experimented with by the industry? To answer these questions, we analyze the impact of digitization—a radical innovation—on business models in the recorded music industry. Using an economic analysis of the effects of digitization, we begin by building five potential digital business models for the music industry. Then, using data from a survey on a sample of French record companies, we map these record labels on our digital business models. Our analysis suggests that digitization has led to a big bang of business models in the music industry, rather than to incremental adjustments of the existing business model.     

Blockchain-based ubiquitous manufacturing: a secure and reliable cyber-physical system

With product customisation and emerging business opportunities, small and medium manufacturing enterprises (SMEs) must find ways to collaborate and share competency in a trustable manner to survive a turbulent market. Therefore, service industry turns to the manufacturing industry and SMEs migrate to cloud manufacturing (CM) and ubiquitous manufacturing. However, existing platforms use centralised networking, which suffers from security, scalability and big-data problems. In this paper, we propose a blockchain-based platform as a trustable network to eradicate third-party problems, which can improve the scalability, security and big-data problems for SMEs. Our proposed platform is developed based on a consortium blockchain which provides a peer-to-peer communication network between the end user and the service provider. We improve existing consensus mechanism and communication protocol based on a cyber-physical system (CPS), via an autonomous agent. Firstly, we provide a review of cloud manufacturing, ubiquitous manufacturing and blockchain-based manufacturing approaches by highlighting the main problems. Then, the proposed platform, blockchain ubiquitous manufacturing (BCUM), is explained, based on its architecture, consensus algorithm and CPS, with the help of autonomous agent communication. The proposed platform has been developed for 3D printing companies which are geographically distributed and tested based on network performance and three practical scenarios.     

Adoption framework for advanced manufacturing technologies

The paper examines the adoption of advanced manufacturing technology (AMT). The general proposition is that the adoption of AMT reflects an interactive learning process, which is described by the extant use of AMT, firm-specific AMT knowledge accumulation, and the transfer of AMT ideas among networks of firms, suppliers, industry associations and government. This general proposition was specified as a structural equation model of Adoption of AMT. Data were based on the experiences of 224 Australian manufacturing companies that have invested and succeeded in using AMT. The results suggest that the adoption of AMT is primarily a function of a firm's assets for AMT, which comprise the resources, capabilities, and competences for developing and using AMT. The relationship between the adoption of AMT and assets for AMT was reciprocal, which implied a process of dynamic accumulation of both technology and human assets — a view consonant with the competence perspective of the firm. Overall, assets for AMT represent the AMT knowledge base of firms, and were shown to have a crucial role in determining the adoption of AMT.     

云计算时代电影工业化体系构建策略与工程设计思考

本文以新时代电影产业提质升级和高质量发展的目标与需求为背景,细致阐述了云计算时代电影工业化体系构建的基本原则、实施策略和重大工程设计,涵盖电影云制作技术服务平台建设工程、电影大数据深度挖掘与智能分析平台建设工程以及基于智能科学技术的智慧电影和智慧影院建设工程,并提出以中国电影高质量发展支撑电影大国向电影强国实现历史性跨越的战略思考。     

计算机关键技术在烟草机械协同并行研发设计中的应用

以中国烟草机械集团有限责任公司及控股企业异地协同并行研发设计平台系统(以下简称协同平台)为例,阐述了计算机及网络的复制、虚拟交换、负载均衡、网闸、CA认证等有关技术在异地协同并行研发设计中的应用.重点介绍PDM系统中的远程数据复制技术、交换机设备的虚拟交换架构,并以应用服务器的负载均衡为例论证了负载均衡技术,从网络、数...     

The role of crowdsourcing in product design: The moderating effect of user expertise and network connectivity

To meet the challenges of customization and personalized design in the context of the digital economy, increasing numbers of small and medium-sized enterprises (SMEs) are using crowdsourcing platforms to seek product design innovation. While crowdsourcing has significant potential for eliciting product ideas, it remains unclear whether it can generate high-quality designs for relatively more complex product design tasks. Based on a leading social new product development website Quirky.com, we explore the antecedents of a firm's adoption decision of crowdsourcing, the impact of crowdsourcing on product design quality, and the moderating effect of user expertise and network connectivity by utilizing treatment effect model. The results indicate that: (1) crowdsourcing has a significant positive impact on product design quality; (2) user expertise positively moderates the role of crowdsourcing in product design quality; (3) network connectivity has an inverted U-shaped moderating effect on crowdsourcing and product design quality relationship; (4) the probability of firm's crowdsourcing adoption is negatively correlated with the complexity and reliability of initial product idea but positively correlated with popularity, novelty, and numbers of images. This study not only enriches the crowdsourcing and social network theory; but also provides guidance for crowdsourcing platform firms to better manage the crowdsourcing process and user crowdsourcing communities.     

Exploring self-organization and self-adaption for smart manufacturing complex networks

Trends toward the globalization of the manufacturing industry and the increasing demands for small-batch, short-cycle, and highly customized products result in complexities and fluctuations in both external and internal manufacturing environments, which poses great challenges to manufacturing enterprises. Fortunately, recent advances in the Industrial Internet of Things (IIoT) and the widespread use of embedded processors and sensors in factories enable collecting real-time manufacturing status data and building cyber–physical systems for smart, flexible, and resilient manufacturing systems. In this context, this paper investigates the mechanisms and methodology of self-organization and self-adaption to tackle exceptions and disturbances in discrete manufacturing processes. Specifically, a general model of smart manufacturing complex networks is constructed using scale-free networks to interconnect heterogeneous manufacturing resources represented by network vertices at multiple levels. Moreover, the capabilities of physical manufacturing resources are encapsulated into virtual manufacturing services using cloud technology, which can be added to or removed from the networks in a plug-and-play manner. Materials, information, and financial assets are passed through interactive links across the networks. Subsequently, analytical target cascading is used to formulate the processes of self-organizing optimal configuration and self-adaptive collaborative control for multilevel key manufacturing resources while particle swarm optimization is used to solve local problems on network vertices. Consequently, an industrial case based on a Chinese engine factory demonstrates the feasibility and efficiency of the proposed model and method in handling typical exceptions. The simulation results show that the proposed mechanism and method outperform the event-triggered rescheduling method, reducing manufacturing cost, manufacturing time, waiting time, and energy consumption, with reasonable computational time. This work potentially enables managers and practitioners to implement active perception, active response, self-organization, and self-adaption solutions in discrete manufacturing enterprises.