User Tools

Site Tools


archives:chinasys-5

5nd ChinaSys Workshop

Oct 17th

13:30~13:40 Opening and Program Overview (Chair: Hong An)

13:40~14:40 Future Directions for CUDA, Junjie Lai (NVIDIA) (Chair: Li Shen)

14:55~16:15 Session 1: EFFICIENT SYSTEM/PLATFORM (Chair: Lintao Zhang)

  • xCloudified: a library OS for cloud service. Zhenyu Guo (郭振宇), MSRA
  • Minerva: a efficient and scalable deep learning training platform. Minjie Wang (王敏捷), MSRA
  • Improving MapReduce performance using smart speculative execution strategy. Zhen Xiao (肖臻), PKU
  • Seraph: an efficient system for parallel processing on a shared graph. Zhi Yang (杨智), PKU

16:30~17:50 Session 2: MEMORY MANAGEMENT (Chair: Haibo Chen)

  • Unordered List ItemSlimDroid: Caching More Apps with Less Memory for Mobile Devices. Weichao Guo (郭维超), Tsinghua
  • Tiling Operators in Stream Programs for Memory Reduction. Chencheng Ye (叶晨成), HUST
  • Pipelined-MapReduce based on Deterministic Producer-Consumer Virtual Memory Model. Huifang Cao (曹慧芳), USTC
  • Rethinking Memory Management for Efficient Hybrid Memory. Dejun Jiang (蒋德钧), ICT

Oct 18th

8:30~9:30 Internet of Things – A computer architecture evolution perspective, Xinming Liu (刘新铭), Intel (Chair: Hong An)

9:45~10:45 Session 3: BIG DATA (Chair: Wenguang Chen)

  • Study on the big data algorithm: a new progressive algorithm for multiple longest common subsequences problem and its efficient parallelization. Yun Xu (徐云), USTC
  • Efficient KV store design and implementatioin on multi-channel SSDs. Guangyu Sun (孙广宇), PKU
  • logKV: 使用Key-Value Store 支持大规模日志处理. Shimin Chen (陈世敏),ICT

11:00~12:00 Session 4: DATA CENTER (Chair: Zhen Xiao)

  • UStore: a low cost cold data storage system for Data Centers. Lintao Zhang (张霖涛), MSRA
  • Dandelion: a practical freely-expandable Data Center network. Binzhang Fu (付斌章), ICT
  • Software defined hardware in Baidu's super scale Data Center. Jian Ouyang (欧阳剑), Baidu

14:00~15:20 Session 5: MISCELLANEOUS (Chair: Yungang Bao)

  • 神经网络计算机. Yunji Chen (陈云霁), ICT
  • XenGT: high performance GPU virtualization with mediated pass-through. Kevin Tian (田坤), Intel
  • MPIzip: 利用程序结构有效地压缩和重放MPI 程序通信Traces. Jidong Zhai (翟季冬), Tsinghua
  • A case for programmable architecture of resourcing on-demand (PARD). Jiuyue Ma (马久跃), ICT

15:40~17:00 Session 6: SIMULATOR & HETEROGENEOUS SYSTEM (Chair: Weihua Zhang)

  • AnaSim: a parallel analytical model based many-core full-system simulator. Haojun Wang (王浩骏), Fudan
  • FAGS:一种快速GPU 模拟器. Cheng Qian (钱程), NUDT
  • iCHAT: Inter-Cache Hardware-Assistant Data Transfer for Heterogeneous Chip Multiprocessors. Junli Gu (谷俊丽), AMD
  • Fair scheduling on dynamic heterogeneous chip multiprocessor. Tao Wang (王涛), USTC

17:00~17:10 Close & Discussion of Next ChinaSys

archives/chinasys-5.txt · Last modified: 2016/06/19 17:39 by yuzhang