CCF YOCSEF深圳将举办计算机系统结构与大数据专题研讨会

您的位置: 首页 > 分论坛 > 深圳

CCF YOCSEF深圳将举办计算机系统结构与大数据专题研讨会

阅读量：459

收藏本文

中国计算机学会青年计算机科技论坛

CCF Young Computer Scientists & Engineers Forum

CCF YOCSEF深圳

时间：2016年12月21日（星期三）14:00-17:30

地点：深圳市南山区白石路深圳大学南校区计算机与软件学院大楼938室

研讨会主题

计算机系统结构与大数据专题研讨会

主办单位：

中国计算机学会青年计算机科技论坛深圳分论坛(CCF YOCSEF深圳)

深圳大学广东省普及型高性能计算机重点实验室

深圳市电脑学会

执行主席：

毛睿 CCF YOCSEF深圳2016-2017 主席

王毅 CCF YOCSEF深圳2016-2017 AC委员、学术秘书

议程

14:00 签到

14:20 报告会开始

CCF YOCSEF深圳组织方致辞

14:30 特邀讲者：黄铠教授，美国南加州大学

演讲题目：云端大数据分析与深度学习在认知服务业的应用

15:30 特邀讲者：孙贤和教授，美国伊利诺伊理工大学

演讲题目：Memory Sluice Gate Theory: Have we found a solution for memory wall?

16:30 特邀讲者：陈子忠教授，美国加州大学河滨分校

演讲题目：面向百亿亿次超级计算机的算法级容错技术

计算机系统结构与大数据专题研讨会

报告1：云端大数据分析与深度学习在认知服务业的应用

特邀讲者：黄铠教授

Kai Hwang is a Professor of EE/CS at the Univ. of Southern California. He received the Ph.D. from UC Berkeley. He has published extensively in computer architecture, parallel processing, cloud computing, and network security. His latest two books, entitled Cloud and Cognitive Computing: A Machine Learning Approach (MIT Press) and Big Data Analytics for Cloud/IoT and Cognitive Learning (Wiley, U.K.) are in press to appear in 2017. An IEEE Life Fellow, he received the very-first CFC Outstanding Achievement Award in 2004 and the Lifetime Achievement Award from IEEE Cloud2012 for his pioneering work in parallel computing and distributed systems. Four of his graduated Ph.D. students were elected as IEEE Fellows and one an IBM Fellow. He has delivered four dozens of keynote or distinguished lectures in international Conferences or Research Centers. Dr. Hwang has performed consulting work with IBM, Intel, MIT Lincoln Lab, Caltech JPL, Chinese Academy of Sciences, and INRIA in France.

报告提要： In this talk, Dr. Hwang addresses the pervasive use of big-data analytics on smart clouds, social networks, intelligent robots, and IoT platforms. He will assess machine/deep learning models and available software tools to advance the cognitive service industry represented by Google, Apple, Nvidia, Baidu, Intel, and IBM, Huawei, etc. The ultimate goal is to achieve enhanced agility, mobility, security, and scalability of public clouds, IoT platforms, and popular social-media networks.
His talk assesses current AI programs and brain projects pursued by high-tech companies, including Google X-Lab, TensorFlow, DeepMind AlphaGo, Nvidia Digits 5 for using GPU in deep learning, IBM neuromorphic computer, and CAS/ICT Camericon project, etc. Some hidden R/D opportunities are revealed for building smart machines，delivery drones, self-driving cars, blockchains, AR/VR gears, and mobile cloud of everythings.

报告2：Memory Sluice Gate Theory: Have we found a solution for memory wall?

特邀讲者：孙贤和教授

Dr. Xian-He Sun received his B.S. in Mathematics from the Beijing Normal University in 1982, and his Master in Mathematics and Ph.D. in Computer Science in 1985 and 1990, respectively, from Michigan State University. He is a University Distinguished Professor of Computer Science at the Illinois Institute of Technology (IIT). He is the director of the Scalable Computing Software laboratory at IIT and a guest faculty in the Mathematics and Computer Science Division at the Argonne National Laboratory. Before joining IIT, he worked at DoE Ames National Laboratory, at ICASE, NASA Langley Research Center, at Louisiana State University, Baton Rouge, and was an ASEE fellow at Navy Research Laboratories. Dr. Sun is an IEEE fellow and is known for his memory-bounded speedup model, also called Sun-Ni’s Law, for scalable computing. His research interests include data-intensive high performance computing, memory and I/O systems, software system for big data applications, and performance evaluation and optimization. He has over 200 publications and 5 patents in these areas. He is a former IEEE CS distinguished speaker, a former vice chair of the IEEE Technical Committee on Scalable Computing, the past chair of the Computer Science Department at IIT, and is serving and served on the editorial board of most of the leading professional journals in the field of parallel processing. Dr. Sun is a member of the Overseas Expert Advisory Committee of the State Council Overseas Chinese Affairs Office (国务院侨办海外专家咨询委员会委员), a member of the Overseas Expert Board of Chinese Academy of Science (中国科学院海外评审专家）, and a member of the China Thousand Talent Plan (short term) (国家千人计划(短期)入选者).

报告提要： The memory-wall problem is a longstanding issue facing the computing community. Many believe the memory-wall problem can only be solved with new memory technologies that improve memory device hardware performance. In this talk, we introduce a system solution, the memory Sluice Gate Theory, for solving the memory-wall problem. The focus of Sluice Gate Theory is not on hardware peak performance, but the achieved memory stall time. Based on Sluice Gate Theory, a memory system is built to mask the performance gap between CPU and memory devices during the data transfer process. The C-AMAT model is used to calculate the data transfer request/supply ratio at each memory layer (sluice stage) dynamically, and a global control algorithm, named layered performance matching (LPM), is developed to match the data transfer at each memory layer and thus match the overall performance between the CPU and memory system. Data concurrency, overlapping, as well as locality, all play a vital role in the matching and transfer process, and LPM makes a unified systematic optimization possible. Experimental testing is conducted which confirm the theoretical findings, with a 150 times performance improvement and the elimination of memory delay impact in our case studies. We will present the concept of Sluice-Gate, the design of C-AMAT and LPM, and discuss the applications and considerations of Sluice-Gate Theory. We will also present some Sluice-Gate-based software solutions for data intensive computing.

报告3：面向百亿亿次超级计算机的算法级容错技术

特邀讲者：陈子忠教授

陈子忠教授，美国加州大学河滨分校超级计算实验室主任、博士生导师、终身教授、美国国家自然科学基金委杰出青年教授奖获得者（U.S. NSF CAREER Award）、Associate Editor for IEEE Transactions on Parallel and Distributed Systems、Subject Area Editor for Elsevier Parallel Computing Journal、国际电气和电子工程师协会（IEEE）高级会员、美国计算机协会（ACM）终身会员。他长期致力于超级计算，云计算，及大数据处理的研究，并以唯一作者在SC、PPoPP、和HPDC上发表论文多篇。

报告提要：下一代百亿亿次超级计算机将会拥有上亿颗计算核心。计算和存储错误在如此庞大的超算系统中不可避免。本报告主要讨论加州大学河滨分校在百亿亿次超级计算机错误恢复机制研究中的最新进展。

参加人员：IT领域专业人士、研究生、媒体、其他有兴趣者

报名联系：王毅 Email：yiwang@szu.edu.cn

参会回执