| Thursday June 12th |
Session 4A: Online Learning I - Theater - 10.45 am ~ 12.30 pm
Session Chair: Lili Su
|
-
Asynchronous Multi-Agent Bandits: Fully Distributed vs. Leader-Coordinated Algorithms
[paper] [abstract]
[3-minute video]
Xuchuang Wang, Yu-Zhen Chen (University of Massachusetts Amherst); Xutong Liu (Carnegie Mellon University); Lin Yang (Nanjing Universary); Mohammad Hajiesmaili (UMass Amherst); Don Towsley (University of Massachusetts - Amherst); John C.S. Lui (The Chinese University of Hong Kong)
-
 Combinatorial Logistic Bandits (Best Paper Finalist)
[abstract]
[3-minute video]
Xutong Liu, Xiangxiang Dai (The Chinese University of Hong Kong); Xuchuang Wang, Mohammad Hajiesmaili (UMass Amherst); John C.S. Lui (The Chinese University of Hong Kong)
-
Online Fair Allocation of Reusable Resources
[paper] [abstract]
Qingsong Liu, Mohammad Hajiesmaili (University of Massachusetts Amherst)
-
Universal and Tight Bounds on Counting Errors of Count-Min Sketch with Conservative Updates
[paper] [abstract]
Younes Ben Mazziane (Inria, University of Avignon LIA); Othmane Marfoq (Meta)
|
Session 4B: Performance I - Lecture Hall 1 - 10.45 am ~ 12.30 pm
Session Chair: Lingda Li
|
-
Internet Service Usage and Delivery As Seen From a Residential Network
[paper] [abstract]
Shuyue Yu, Thomas Koch, Ilgar Mammadov, Hangpu Cao, Gil Zussman, Ethan Katz-Bassett (Columbia University)
-
The Last Survivor of PoS Pools: Staker's Dilemma
[paper] [abstract]
[3-minute video]
Yuming Huang (National University of Singapore); Jing Tang (The Hong Kong University of Science and Technology); Qianhao Cong, Richard Ma (National University of Singapore); Lei Chen (Hong Kong University of Science and Technology); Yeow Meng Chee (National University of Singapore)
-
Understanding Intel User Interrupts
[paper] [abstract]
Yves Kone (IRIT, Université de Toulouse, CNRS, Toulouse INP, UT3 Toulouse); Louis Duval, Renaud Lachaize (Univ. Grenoble Alpes, CNRS, Inria, Grenoble INP, LIG); Pascal Felber (University of Neuchâtel); Daniel Hagimont (IRIT, Université de Toulouse, CNRS, Toulouse INP, UT3 Toulouse); Alain Tchana (Univ. Grenoble Alpes, CNRS, Inria, Grenoble INP, LIG)
-
A Case Study for Ray Tracing Cores: Performance Insights with Breadth-First Search and Triangle Counting in Graphs
[paper] [abstract]
Zhixiong Xiao, Mengbai Xiao, Yuan Yuan, Dongxiao Yu (Shandong University); Rubao Lee (Freelance); Xiaodong Zhang (The Ohio State University)
|
Session 4C: Quantum - Lecture Hall 2 - 10.45 am ~ 12.30 pm
Session Chair: Sushil Varma
|
-
Peer-to-Peer Distribution of Graph States Across Spacetime Quantum Networks of Arbitrary Topology
[paper] [abstract]
Yuexun Huang (The University of Chicago); Xiangyu Ren (The University of Edinburgh); Bikun Li, Yat Wong (The University of Chicago); Zhiding Liang (Rensselaer Polytechnic Institute); Liang Jiang (The University of Chicago)
-
Modeling and Simulating Rydberg Atom Quantum Computers for Hardware-Software Co-design with PachinQo
[paper] [abstract]
[3-minute video]
Jason Ludmir, Yuqian Huo, Nicholas S. DiBrita, Tirthak Patel (Rice University)
-
Optimal Scheduling in a Quantum Switch: Capacity and Throughput Optimality
[abstract]
Sanidhay Bhambay, Thirupathaiah Vasantam, Neil Walton (Durham University)
-
Quantum Network Optimization: From Optimal Routing to Fair Resource Allocation
[paper] [abstract]
Zhaozhen Wang, Xingang Shi, Zhengfeng Ji, Xia Yin (Tsinghua University)
|
Rising Star Award: Debankur Mukherjee
Mean-Field Methods for Constrained Systems: Revisiting Load Balancing Under Data Locality
Theater - 1.30 pm ~ 2.30 pm
Session Chair: Atilla Eryilmaz
|
-
Abstract: Modern data centers and cloud platforms rely heavily on efficient load balancing algorithms to meet stringent performance targets. A key challenge in this setting is data locality constraints that determine which tasks can be processed by which servers, often modeled as a bipartite compatibility graph. These constraints break the exchangeability of servers typically assumed in classical models, undermining the validity of standard mean-field approximations and causing a gap between theory and practice.
In this talk, I will present a sequence of results that extend the scope of mean-field analysis to systems with data locality constraints. I will discuss how the structural properties of the underlying graph, whether it is an expander, a spatial graph, or highly inhomogeneous, can fundamentally alter system dynamics. Along the way, I will describe a framework that allows us to derive performance guarantees and asymptotically optimal designs, even under strong constraints. This line of work leverages stochastic coupling techniques with novel notions of graph expansion and mixing time ideas, and points toward new principles for designing scalable, resource-efficient systems under structural limitations.
-
Bio: Debankur Mukherjee is the Leo and Louise Benatar Early Career Professor and Assistant Professor in the H. Milton Stewart School of Industrial and Systems Engineering at the Georgia Institute of Technology. Before joining Georgia Tech in 2019, he was a Prager assistant professor for a year in the Division of Applied Mathematics at Brown University. Debankur got his Ph.D. in Stochastic Operations Research from the Eindhoven University of Technology in the Netherlands. Debankur’s research spans the area of applied probability, at the interface of stochastic processes and computer science, with applications to performance analysis, online algorithms, and machine learning. His primary focus is to develop a foundational understanding of the challenges that arise in large-scale systems, such as data centers and cloud networks. His work was a finalist in the INFORMS JFIG paper competition in 2022 and, received the Best Paper Award at ACM SIGMETRICS 2023 and the Best Student Paper Award at ACM SIGMETRICS 2018. His research has been funded by the NSF and he is currently serving on the editorial boards of Stochastic Systems, QUESTA, and Stochastic Models.
|
Session 5A: Systems I - Theater - 2.30 pm ~ 3.30 pm
Session Chair: Benny Van Houdt
|
-
NetJIT: Bridging the Gap from Traffic Prediction to Preknowledge for Distributed Machine Learning
[paper] [abstract]
Xin Ai, Zijian Li, Yuanyi Zhu, Zixuan Chen, Sen Liu, Yang Xu (Fudan University)
-
ScaleOPT: A Scalable Optimal Page Replacement Policy Simulator
[paper] [abstract]
Hyungseok Han (goorm Inc.); Sangjin Lee, Yongseok Son (Chung-Ang University)
|
Session 5B: Systems II - Lecture Hall 1 - 2.30 pm ~ 3.30 pm
Session Chair: Vijay Subramanian
|
-
PipeCo: Pipelining Cold Start of Deep Learning Inference Services on Serverless Platforms
[paper] [abstract]
Jiaang Duan, Shiyou Qian, Hanwen Hu (Shanghai Jiao Tong University); Dingyu Yang (Zhejiang University); Jian Cao (Shanghai Jiaotong University); Guangtao Xue (Shanghai Jiao Tong University)
-
PyGim: An Efficient Graph Neural Network Library for Real Processing-In-Memory Architectures
[paper] [abstract]
Christina Giannoula, Peiming Yang (University of Toronto); Ivan Fernandez (Barcelona Supercomputing Center); Jiacheng Yang, Sankeerth Durvasula, Yu Xin Li (University of Toronto); Mohammad Sadrosadati (ETH Zürich); Juan Gomez Luna (NVIDIA); Onur Mutlu (ETH Zurich and Stanford University); Gennady Pekhimenko (University of Toronto / Vector Institute)
|
Session 5C: Performance II - Lecture Hall 2 - 2.30 pm ~ 3.30 pm
Session Chair: Ziv Scully
|
-
CertainSync: Rateless Set Reconciliation with Certainty
[paper] [abstract]
Tomer Keniagin, Eitan Yaakobi, Ori Rottenstreich (Technion - Israel Institute of Technology)
-
CHash: A High Cost-Performance Hash Design for CXL-based Disaggregated Memory System
[paper] [abstract]
Mengting Lu, Gaocong Liu, Kun Wang, Feng Zhu, Shu Li (Alibaba Group)
|
Session 6A: Online learning II - Theater - 4.00 pm ~ 5.45 pm
Session Chair: Weina Wang
|
-
Learning-Augmented Decentralized Online Convex Optimization in Networks
[paper] [abstract]
[3-minute video]
Pengfei Li (UC Riverside); Jianyi Yang (University of Houston); Adam Wierman (Caltech); Shaolei Ren (UC Riverside)
-
Robust Gittins for Stochastic Scheduling
[abstract]
Benjamin Moseley, Heather Newman (Carnegie Mellon University); Kirk Pruhs (University of Pittsburgh); Rudy Zhou (Microsoft)
-
Learning-Augmented Competitive Algorithms for Spatiotemporal Online Allocation with Deadline Constraints
[paper] [abstract]
[3-minute video]
Adam Lechowicz (University of Massachusetts Amherst); Nicolas Christianson (California Institute of Technology); Bo Sun (University of Waterloo); Noman Bashir (Massachusetts Institute of Technology); Mohammad Hajiesmaili (University of Massachusetts Amherst); Adam Wierman (California Institute of Technology); Prashant Shenoy (University of Massachusetts Amherst)
-
A Gittins Policy for Optimizing Tail Latency
[paper] [abstract]
Amit Harlev, George Yu, Ziv Scully (Cornell University)
|
Session 6B: Measurement II - Lecture Hall 1 - 4.00 pm ~ 5.45 pm
Session Chair: Timothy Zhu
|
-
Revisiting Traffic Splitting for Software Switch in Datacenter
[paper] [abstract]
[3-minute video]
Yeonho Yoo, Gyeongsik Yang, Changyong Shin, Hwiju Cho, Wonmi Choi (Korea University); Zhixiong Niu (Microsoft Research); Chuck Yoo (Korea University)
-
Exploiting Kubernetes Autoscaling for Economic Denial of Sustainability
[paper] [abstract]
[3-minute video]
Jonathan Chamberlain, Jilin Zheng (Boston University); Zeying Zhu, Zaoxing Liu (University of Maryland); David Starobinski (Boston University)
-
A Global Perspective on the Past, Present, and Future of Video Streaming over Starlink
[paper] [abstract]
Liz Izhikevich (University of California, Los Angeles); Reese Enghardt, Te-Yuan Huang, Renata Teixeira (Netflix)
-
ForgetMeNot: Understanding and Modeling the Impact of Forever Chemicals Toward Sustainable Large-Scale Computing
[paper] [abstract]
[3-minute video]
Rohan Basu Roy (University of Utah); Raghavendra Kanakagiri (Indian Institute of Technology Tirupati); Yankai Jiang, Devesh Tiwari (Northeastern University)
|
Session 6C: Blockchains - Lecture Hall 2 - 4.00 pm ~ 5.45 pm
Session Chair: Florin Ciucu
|
-
Phishing Tactics Are Evolving: An Empirical Study of Phishing Contracts on Ethereum
[paper] [abstract]
Bowen He (Zhejiang University); Xiaohui Hu (Huazhong University of Science and Technology); Yufeng Hu (Zhejiang University); Ting Yu (Mohamed bin Zayed University of Artificial Intelligence); Rui Chang, Lei Wu, Yajin Zhou (Zhejiang University)
-
Blockchain Amplification Attack
[paper] [abstract]
Taro Tsuchiya (Carnegie Mellon University); Liyi Zhou (The University of Sydney); Kaihua Qin (Yale University); Arthur Gervais (University College London & UC Berkeley RDI); Nicolas Christin (Carnegie Mellon University)
-
Piecing Together the Jigsaw Puzzle of Transactions on Heterogeneous Blockchain Networks
[paper] [abstract]
[3-minute video]
Xiaohui Hu (Huazhong University of Science and Technology); Hang Feng (Zhejiang University); Pengcheng Xia (Beijing University of Posts and Telecommunications); Gareth Tyson (Hong Kong University of Science and Technology (GZ)); Lei Wu, Yajin Zhou (Zhejiang University); Haoyu Wang (Huazhong University of Science and Technology)
-
Towards Understanding and Analyzing Instant Cryptocurrency Exchanges
[paper] [abstract]
Yufeng Hu, Yingshi Sun, Lei Wu, Yajin Zhou, Rui Chang (Zhejiang University)
|
