icloud-ecnu / paper-reading-list

Reading paper list for iCloud group

paper-reading-list

Reading paper list for iCloud-ECNU Group

research directions:

Performance-centric Cloud Interface (Predictable Performance with Cost-Efficient Resource Provisioning for Cloud Users)

HarmonyBatch: Batching multi-SLO DNN inference with heterogeneous serverless functions (Jiabin-June 2024, IWQoS'24)
spotDNN: Spot instance provisioning for predictable distributed DNN training (Ruitao-June 2023, IWQoS'23)
LambdaDNN: Serverless function provisioning for predictable distributed DNN training (Yiling-June 2021, TC'22)
Cynthia: VM instance provisioning for predictable distributed DNN training (Haoyue-June 2020, ICPP'19 & TPDS'19)
iSpot: Spot instance provisioning for predictable big data analytics (Huan-June 2019, ISPA'17)

MLSys (Co-optimization of DL Job Scheduling and Cluster Resource Allocation)

Performance optimization of LLM inference on multi-engine systems (Zhuoyan)
Reconfigurable job scheduling in multi-tenant deep learning clusters (Xinyi)
Predictable performance of LLM training in heterogeneous clusters (Qiannan)
Opara: Speeding up DNN inference with inter-operator parallelism (Aodong-June 2024, CCFSys'23 & TC'24)
Nebula & Prophet: Communication scheduling optmization for distributed machine learning (Qiang-June 2021, ICPP'21 & THPC'21)
DelayStage: Tasks scheduling for performance optimization of DAG-style data analytics (Wujie-June 2020, ICPP'19)

Interference-aware Cluster Resource Management (Container Scheduling and GPU Resource Management)

U²CMig: User-unaware live migration of containers with memory dirty page prediction (Shuohao-June 2024)
Tetris: Scheduling long-running containerized workloads for load balancing in shared clusters (Xiyue-June 2023)
iGniter: Interference-aware GPU resource provisioning for predictable DNN inference (Jianian-June 2022, TPDS'23)

Collaborative Cloud-Edge-Client Computing

eBrowser: Energy-efficient mobile Web browser (Shuai-June 2018, ICDCS'18)