Myeongjae Jeon

Assistant professor @ ECE UNIST
mjjeon_at_unist.ac.kr
[CV]

I have openings for the following positions. Please contact me by email, if interested.

1) A post-doc researcher in large-scale real-time data analytics. The primary role will be leading a collaborative research with Microsoft Research & Microsoft Azure. 2) Graduate students, and post-master/undergraduate researchers.

I am an assistant professor in the Department of Computer Science and Engineering at Ulsan National Institute of Science and Technology (UNIST) since 2018 fall. Prior to joining UNIST, I was with Systems Research Group in Microsoft Research Redmond (from 2015-2018), and before that I worked at ARM R&D (from 2014-2015). 

I finished my Ph.D. in computer science at Rice University in May 2014, under the supervision of Prof. Alan L. Cox and Prof. Scott Rixner. During my Ph.D. I actively collaborated with Dr. Yuxiong He and Dr. Sameh Elnikety at Microsoft Research. I received the M.S. degree in computer science from KAIST in 2009 and the B.E. degree in computer engineering from Kwangwoon University in 2005.

Research

My research interests span distributed systems, data analytics engines, computer architecture, and applied machine learning. My research goal is to advance the state of the art in emerging large-scale computing platforms by making them more efficient, responsive, intelligent and programmable. At UNIST, I plan to devote myself on the following areas:
  • Parallel/distributed processing of deep learning workloads
  • Real-time stream data analytics at cloud/IoT scale
  • Public/private blockchain
I am also interested in applying my expertise on machine learning for systems into other systems. Here is the summary of the work.
  • Reducing the latency of interactive processing (i.e., search) in big data platform in the context of a commercial search engine, Microsoft Bing.
  • Reducing high-percentile query latency over varying workload characteristics through adaptive resource management algorithms [ASPLOS16] that leverage two key techniques, query parallelism [EuroSys13] and query execution time prediction [SIGIR14].

Publications

  1. Multi-tenant GPU Clusters for Deep Learning Workloads: Analysis and Implications
    • Myeongjae Jeon, Shivaram Venkataraman, Junjie Qian, Amar Phanishayee, Wencong Xiao, Fan Yang
    • MSR-TR, May. 2018 [PDF]

  2. TerseCades: Efficient Data Compression in Stream Processing
    • Gennady Pekhimenko, Chuanxiong Guo, Myeongjae Jeon, Peng Huang, Lidong Zhou
    • USENIX ATC, Jul. 2018 [PDF][Slides]

  3. StreamBox: Modern Stream Processing on a Multicore Machine
    • Hongyu Miao, Heejin Park, Myeongjae Jeon, Gennady Pekhimenko, Kathryn S. McKinley, Felix Xiaozhu Lin
    • USENIX ATC, Jul. 2017 [PDF][Slides]

  4. SSD Failures in Datacenters: What, When and Why?
    • Iyswarya Narayanan, Di Wang, Myeongjae Jeon, Bikash Sharma, Laura Caulfield, Anand Sivasubramaniam, Ben Cutler, Jie Liu, Badriddine Khessib, Kushagra Vaid
    • ACM SIGMETRICS / IFIP Performance, Jun. 2016 [PDF] (Poster)
    • ACM SYSTOR, Jun. 2016 (Best Student Paper) [PDF]

  5. TPC: Target-Driven Parallelism Combining Prediction and Correction to Reduce Tail Latency in Interactive Services
    • Myeongjae Jeon, Yuxiong He, Hwanju Kim, Sameh Elnikety, Scott Rixner, Alan L. Cox
    • ASPLOS, Apr. 2016 [PDF][Slides]

  6. Predictive Parallelization: Taming Tail Latencies in Web Search
    • Myeongjae Jeon, Saehoon Kim, Seung-Won Hwang, Yuxiong He, Sameh Elnikety, Alan L. Cox, Scott Rixner
    • ACM SIGIR, Jul. 2014 [PDF][Slides(pptx)]

  7. Reducing DRAM Row Activations with Eager Read/Write Clustering
    • Myeongjae Jeon, Conglong Li, Alan L. Cox, Scott Rixner
    • ACM Transactions on Architecture and Code Optimization (TACO), Dec. 2013 [PDF]

  8. Adaptive Parallelism for Web Search
    • Myeongjae Jeon, Yuxiong He, Sameh Elnikety, Alan L. Cox, Scott Rixner
    • ACM EuroSys, Apr. 2013 [PDF][Slides(pptx)]

  9. Workload Characterization and Performance Implications of Large-Scale Blog Servers
    • Myeongjae Jeon, Youngjae Kim, Jaeho Hwang, Joonwon Lee, Euiseong Seo
    • ACM Transactions on the Web (TWEB), Nov. 2012 [PDF]

  10. Energy Reduction in Consolidated Servers Through Memory-Aware Virtual Machine Scheduling
    • Jae-Wan Jang, Myeongjae Jeon, Hyo-Sil Kim, Heeseung Jo, Jin-Soo Kim, Seungryoul Maeng
    • IEEE Transactions on Computers (TC), Apr. 2011 [PDF]

  11. Replicated Abstract Data Types: Building Blocks for Collaborative Applications
    • Hyun-Gul Roh, Myeongjae Jeon, Jinsoo Kim, Joonwon Lee
    • Journal of Parallel and Distributed Computing (JPDC), Mar. 2011 [PDF]

  12. Measurement, Modeling, and Analysis of a Large-scale Blog Server Workload
    • Myeongjae Jeon, Jaeho Hwang, Youngjae Kim, Jae-wan Jang, Joonwon Lee, Euiseong Seo
    • IEEE SocialCom SCA, Aug. 2010 [PDF]

  13. Log' version vector: Logging version vectors concisely in dynamic replication
    • Hyun-Gul Roh, Myeongjae Jeon, Euiseong Seo, Jinsoo Kim, Joonwon Lee
    • Elsevier Information Processing Letters (IPL), May. 2010 [PDF]

  14. Guest-Aware Priority-based Virtual Machine Scheduling for Highly Consolidated Server
    • Dongsung Kim, Hwanju Kim, Myeongjae Jeon, Euiseong Seo, Joonwon Lee
    • Euro-Par, Aug. 2008 [PDF]

  15. Domain Level Page Sharing in Xen Virtual Machine Systems
    • Myeongjae Jeon, Euiseong Seo, Junghyun Kim, Joonwon Lee
    • APPT, Nov. 2007 [PDF]

Prior Work Experiences

  • Research Intern, Microsoft Research, Summer 2011, Summer 2012, Fall 2012, Summer 2013
  • Software Design Engineer, Microsoft Research (contracted via Collabera Inc.), Jan. 2013 - May 2014
  • Open Source Developer, HP Labs' COTSon full-system simulation platform, since May 2011