User Tools

Site Tools


program

ACM TURC 2018 SIGOPS (14th ChinaSys) Workshop Program

May 18th - 19th, 2018

Fuyue Hotel (上海富悦大酒店), Shanghai, China

Organized by ACM SIGOPS ChinaSys, USTC

Sponsored by


Day 1: May 18th

Time Talks
08:45 - 09:00 Openning Remarks (Wenguang Chen, Yu Zhang)
09:00 - 10:00 Keynote 1
深度学习处理器
Speaker: Yunji Chen (Professor, ICT)
Chair: Yu Zhang (Associate Professor, USTC)
10:00 - 10:20 Tea Break
10:20 - 12:00 Session 1: Fighting Distributed Environments
Chair: Yubin Xia (Associate Professor, SJTU)
TcpRT: ⼤规模云数据库的实时服务质量分析诊断系统
Yusong Gao (Alibaba Group)
Datasize-Aware High Dimensional Configurations Auto-Tuning of In-Memory Cluster Computing
Zhibin Yu, Zhendong Bei (Shenzhen Institute of Advanced Technology); Xuehai Qian (University of Southern California Los Angeles)
深度学习模型的分布式部署优化
Cheng Yang, Sa Wang, Yungang Bao, Wenfeng Xu (ICT)
X-Paxos: A global distributed consensus protocol library
Yingqiang Zhang (Alibaba Group)
QFrag: Distributed Graph Search via Subgraph Isomorphism
Marco Serafini (QCRI)
12:00 - 14:00 Lunch and Break
14:00 - 15:00 Keynote 2
智能驾驶:一个复杂的系统工程
Speaker: Bo Huang (General Manager, UISEE)
Chair: Xiaosong Ma (Principal Scientist, QCRI)
15:00 - 15:15 Tea Break
15:15 - 16:55 Session 2: OS and GPUs
Chair: Yungang Bao (Professor & Director of ACS, ICT)
Avalon: Building an Operating System for Robotcenter
Yuan Xu, Zhiyuan Yan, Sa Wang, Cheng Yang, Qingsai Xiao, Yungang Bao (ICT)
Get Out of the Valley: Power-Efficient Address Mapping for GPUs
Yuxi Liu (Ghent University & Peking University); Xia Zhao (Ghent University); Magnus Jahre (Norwegian University of Science and Technology); Zhenlin Wang (Michigan Technological University); Xiaolin Wang, Yingwei Luo (Peking University); Lieven Eeckhout (Ghent University) [slides]
Transparent Partial Page Migration between CPU and GPU
Shiqing Zhang, Yaohua Yang, Li Shen, Zhiying Wang (NUDT)
DCUDA: Dynamic GPU Sharing via Live Migration
Fan Guo (USTC); Junping Zhao (Dell EMC); Yongkun Li (USTC); Kun Wang (Dell EMC); Yinlong Xu (USTC); John C. S. Lui (CUHK)
Accelerating the GPU Database Operations via a Data-driven Approach
Rentong Guo, Chao Xie (Zilliz); Xiaofei Liao (HUST)
16:55 - 17:10 Tea Break
17:10 - 18:00 Panel: ML system builder and user, both hardware and software
Moderator: Zheng Zhang (Professor, NYU)
Panelist:
Yiran Chen (Associate Professor & Co-Director of CEI, Duke University)
Yu Wang (Associate Professor, Tsinghua University / Co-Founder, Deephi)
Bo Huang (General Manager, UISEE)
Naiyan Wang (Co-Founder & Chief Scientist, TuSimple)
Haibo Chen (Professor, SJTU / Chief Scientist for OS, Huawei Research)

Day 2: May 19th

Time Talks
14:00 - 15:00 Keynote 3
Designing An Efficient and Safe Neuromorphic Computing System
Speaker: Yiran Chen (Associate Professor & Co-Director of CEI, Duke University)
Chair: Wenguang Chen (Professor, Tsinghua University)
15:00 - 15:15 Tea Break
15:15 - 16:35 Session 3: From Cache to DFS
Chair: Haikun Liu (Associate Professor, HUST)
Penalty-Aware Cache Modeling and Its Applications
Cheng Pan, Xiameng Hu, Lan Zhou, Yingwei Luo, Xiaolin Wang (Peking University); Zhenlin Wang (Michigan Tech)
UKSM: Swift Memory Deduplication via Hierarchical and Adaptive Memory Region Distilling
Nai Xia, Chen Tian (Nanjing University); Yan Luo, Hang Liu (University of Massachusetts Lowell); Xiaoliang Wang (Nanjing University)
RAID+: Deterministic and Balanced Data Distribution for Large Disk Enclosures
Guangyan Zhang, Zican Huang (Tsinghua University); Xiaosong Ma (QCRI, HBKU); Songlin Yang, Zhufan Wang, Weimin Zheng (Tsinghua University)
Improving Metadata Performance in Distributed File Systems via Data Correlation-Directed Prefetching
Youxu Chen, Cheng Li, Min Lyu, Xinyang Shao, Yongkun Li, Yinlong Xu (USTC)
16:35 - 17:35 Session 4: Virtual is A Power
Chair: Lei Wang, Associate (Professor, NUDT)
VButton: Practical Attestation of User-driven Operations in Mobile Apps
Wenhao Li, Shiyu Luo, Zhichuang Sun, Yubin Xia, Long Lu, Haibo Chen, Binyu Zang, Haibing Guan (SJTU)
When DBT meets JIT
Xiaoli Gong, Tao Li (Nankai University)
EPTI: Efficient Defence against Meltdown Attack for Unpatched VMs
Zhichao Hua, Dong Du, Yubin Xia, Haibo Chen, Binyu Zang (SJTU) [slides]
17:35 - 17:50 Concluding Remarks

Keynotes

Bo Huang

黄波

简介:黄波博士,毕业于复旦大学计算机科学系计算机软件与理论专业,驭势(上海)汽车科技有限公司总经理。加盟驭势前一直在英特尔亚太研发有限公司工作,是英特尔中国本土培养的第一位资深首席工程师。在英特尔工作期间,他创建了英特尔中国编译技术组并带领团队开发了多款编译器和动态优化器。他还倡导端到端的系统优化并在云端的服务器以及客户端的软件与服务优化上构建了英特尔中国的核心竞争力。作为技术创新的推动者,黄波博士希望用科技来改变人类的生活。

题目智能驾驶:一个复杂的系统工程

摘要:智能驾驶是未来10年人工智能的重要应用之一。在本演讲中,演讲者将分享一个典型的智能驾驶系统的系统架构,系统中各模块的主要功能以及开发一个产品级的智能驾驶系统所需要考虑的一些问题。



Yunji Chen

陈云霁

简介:陈云霁,中国科学院计算技术研究所研究员,博士生导师。同时,他担任了中国科学院脑科学卓越中心特聘研究员,以及中国科学院大学岗位教授。目前他带领其实验室,研制寒武纪系列深度学习处理器。在此之前,他从事国产处理器的研发工作十余年,先后负责或参与了多款龙芯处理器的设计。他在包括ISCA、HPCA、MICRO、ASPLOS、ICSE、ISSCC、Hot Chips、IJCAI、FPGA、SPAA、IEEE Micro以及8种IEEE/ACM Trans.在内的学术会议及期刊上发表论文60余篇。陈云霁获得了首届国家自然科学基金“优秀青年基金”、首届国家万人计划“青年拔尖人才”、中国计算机学会青年科学家奖以及中科院青年人才奖。他还作为负责人带领科研团队获得了全国“青年文明号”和中央国家机关“青年文明号”的称号。

题目:深度学习处理器

摘要:以深度学习为代表的人工神经网络是机器学习最重要的方法之一,在云端和终端都有非常广泛的应用(例如广告推荐、自动翻译、语音识别、图像识别等)。然而传统的 CPU 和 GPU 芯片在进行神经网络处理时遇到了严重的性能和能耗瓶颈。近年来,我们和 Inria 合作设计了国际上首个深度学习处理器架构,能将深度学习处理能耗降低多个数量级。相关工作获得了 ASPLOS'14 和 MICRO'14 的最佳论文奖(亚洲迄今仅有的两次获计算机体系结构顶级国际会议最佳论文),并入选了 CACM 评选的研究亮点,引起了国际同行的广泛关注。我们研制的寒武纪深度学习处理器已经应用在包括华为Mate10手机在内的千万用户终端中。



Yiran Chen

陈怡然

简介:Yiran Chen received B.S and M.S. from Tsinghua University and Ph.D. from Purdue University in 2005. After five years in industry, he joined University of Pittsburgh in 2010 as Assistant Professor and then promoted to Associate Professor with tenure in 2014, held Bicentennial Alumni Faculty Fellow. He now is a tenured Associate Professor of the Department of Electrical and Computer Engineering at Duke University and serving as the co-director of Duke Center for Evolutionary Intelligence (CEI), focusing on the research of new memory and storage systems, machine learning and neuromorphic computing, and mobile computing systems. Dr. Chen has published one book and more than 300 technical publications and has been granted 93 US patents. He is the associate editor of IEEE TNNLS, IEEE TCAD, IEEE D&T, IEEE ESL, ACM JETC, ACM TCPS, and served on the technical and organization committees of more than 40 international conferences. He received 6 best paper awards and 14 best paper nominations from international conferences. He is the recipient of NSF CAREER award and ACM SIGDA outstanding new faculty award. He is the Fellow of IEEE.

题目: Designing An Efficient and Safe Neuromorphic Computing System

摘要: Neuromorphic computing was originally referred to as the hardware that mimics neuro-biological architectures to implement models of neural systems. The concept was then extended to the computing systems that can run bio-inspired computing models, e.g., neural networks and deep learning networks. As an alternative computing platform targeting cognitive applications, power efficiency is one of the most critical design metrics of neuromorphic computing systems. The computational efficiency of a neuromorphic computing system can be improved through innovations in both computing models and hardware designs, which also generate many interesting interactive optimizations. Very recently, the security of neural network models and the consumed data emerge as another crucial concern of neuromorphic computing systems. In this talk, we will discuss several technical approaches of designing an efficient and safe neuromorphic computing system.

program.txt · Last modified: 2018/05/25 12:11 by root