Jinlai Xu (徐锦来)

Jinlai Xu (徐锦来)

Researcher/Engineer

Infrastructure System Lab, ByteDance Inc., US

Biography

Jinlai Xu is a Researcher/Engineer at Infrastructure System Lab of ByteDance US. He is currently working on cutting-edge technologies which will evolve the infrastructure to a new generation, including but not limited to Next Generation ML/Big Data infra, Graph Learning/Computing, Cloud Native & Serverless Infra and Hyper-Scale Heterogeneous Cluster Management.

He got his PhD from University of Pittsburgh in 2021. His research interests include Big Model Training/Inference/Finetune Framework, Serverless Computing, Distributed Systems, Fog/Edge and Cloud Computing, Stream Processing Optimization and Blockchain-based Techniques. His PhD advisor is Balaji Palanisamy. He got his bachelor and master from China University of Geosciences and his undergraduate and graduate advisor is Zhongwen Luo.

Full Curriculum Vitae

Interests
  • Big Model
  • Serverless Computing
  • Fog/Edge Computing
  • Cloud Computing
  • Distributed Systems
  • Stream Processing Optimization
  • Reinforcement Learning
  • Blockchain
Education
  • PhD in Infomation Science, 2021

    University of Pittsburgh

  • M.Phil in Software Engineering, 2015

    China University of Geosciences

  • BEng in Software Engineering, 2012

    China University of Geosciences

Research Topics

Big Model

Big Model training/inference/finetune Framework

Serverless Computing

Ray Project

Stream Processing

Low-latency Stream Processing/Resilience/Elasticity in Edge Computing

Edge Computing

Resource Allocation/Management/Sharing

Cloud Computing

Resource Allocation/Management/Sharing

Reinforcement Learning

RL on Systems

Blockchain

Incentive Deisign for Edge Resource Sharing

Teaching Experience

  • Teaching Assistant , September 2015 - Present

    • University of Pittsburgh
      • Cloud Computing (2017 Spring, 2018 Spring, 2019 Spring)
      • Information Security & Privacy (2017 Fall)
      • Information Security & Privacy (Online Course) (2018 Fall)
      • Algorithm Design (2018 Fall)
  • Teaching Assistant , September 2013 - January 2014

    • China University of Geosciences
      • Advanced Programming Language (JAVA)
        • Instructor: Prof. Shengwen Li

Professional Services

Research Experience

 
 
 
 
 
Researcher/Engineer
Infrastructure System Lab, ByteDance US
Jan 2022 – Present Seattle, WA, USA
  • Working on cutting-edge technologies which will evolve the infrastructure to a new generation, including but not limited to Next Generation ML/Big Data infra, Graph Learning/Computing, Cloud Native & Serverless Infra and Hyper-Scale Heterogeneous Cluster Management.
  • Focus on projects related to serverless computing (Ray project), Graph Computing, Big Model training/inference/finetune Framework, etc.
 
 
 
 
 
PhD Student
University of Pittsburgh
Sep 2015 – Dec 2021 Pittsburgh, PA, USA
  • Reviewed related literatures (mainly in Distributed Systems, Cloud Computing, Edge Computing, Stream Processing, Reinforcement Learning and Blockchain-based Techniques)
  • Focus on resource management problems in Edge and Cloud Computing to achieve low-latency stream processing
  • Publish papers on these topics
 
 
 
 
 
Research Assistant
China University of Geosciences
Sep 2012 – Jun 2015 Wuhan, Hubei, China
  • Reviewed related literatures (mainly in Cloud Computing)
  • Constructed the cloud computing platform for our faculty:
    • Designed the virtualization solution for the cluster. (based on Xen)
    • Deployed Hadoop and related application(Hive, Spark, Solr …) on the cluster.
    • Supported the experiment of Deep Learning in our lab.
  • Studied MapReduce programming model and did research on it:
    • Read the source code of MapReduce in Hadoop project.
    • Proposed a new method to reuse the intermediate results automatically and data-awarenessly and implemented the prototype system by modifying the core code of MapReduce.
    • Evaluated the performance on the cluster and got the result that the system could improve the performance up to 24.6% compared with the previous optimization work.
    • The paper is published on CCPE. (Title: MEMoMR: Accelerate MapReduce via reuse of intermediate results)
  • Managed the cluster in our faculty:
 
 
 
 
 
Undergraduate Research Assistant
China University of Geosciences
Jul 2009 – Aug 2012 Wuhan, Hubei, China
  • Reviewed related literatures (mainly in Computer Vision and Robotics).
  • Participated in The 9 th Robot Soccer Tournament of China and The Tryouts for FIRA in Changchun in freshmen year.
  • Studied the architecture and implementation of ROS(The Robot Operating System) and preliminarily deployed it on the robots control panel (Version: RB100 by RoBoard).
  • Successfully applied for The National College Students Innovation Experiment Program (with funding):
    • Topic: Small Model Aircraft Autopilot System and Aerial Photo Research
    • Chose Quadrotor(an aircraft with four rotors) as the carrier platform of the research.
    • Studied the theory of balancing the Quadrotor with MikroKopter(one of the most famous open source UAV projects).
    • Studied and implemented the point clouds registration algorithm ICP and RANSAC on ROS.
    • Used ASUS Xtion PRO (a device like Kinect) to get the point cloud data and evaluated the algorithm.
    • Wrote graduation thesis based on this topic.(Title: the Design and Implementation of the Quadrotor Autopilot and 3-D Point Cloud Generation and Processing System)

Contact