SOSP 2023 Program

Unless otherwise indicated, all sessions will take place at the Rhein-Mosel-Halle.

Monday, October 23, 2023

Workshops 08:30 - 18:00 (UTC+02:00)
Welcome Reception
The welcome reception will happen in Schloss Koblenz (20-minute walk from the main venue [walking directions]). 18:30 - 19:30 (UTC+02:00)

Tuesday, October 24, 2023

Welcome 08:30 - 09:00 (UTC+02:00)
Session 1: Kernel Design and Testing (Chair: Gernot Heiser) 09:00 - 10:30 (UTC+02:00)
Paper TreeSLS: A Whole-system Persistent Microkernel with Tree-structured State Checkpoint on NVM
Best Paper Award
Fangnuo Wu (Institute of Parallel and Distributed Systems, Shanghai Jiao Tong University), Mingkai Dong (Institute of Parallel and Distributed Systems, Shanghai Jiao Tong University), Gequan Mo (Shanghai Jiao Tong University), Haibo Chen (Shanghai Jiao Tong University)
Paper MEMTIS: Efficient Memory Tiering with Dynamic Page Classification and Page Size Determination
Taehyung Lee (Sungkyunkwan University), Sumit Kumar Monga (Virginia Tech), Changwoo Min (Igalia), Young Ik Eom (Dept. of Electrical and Computer Engineering / College of Computing and Informatics, Sungkyunkwan University)
Paper Snowcat: Efficient Kernel Concurrency Testing using a Learned Coverage Predictor
Sishuai Gong (Purdue University), Dinglan Peng (Purdue University), Deniz Altınbüken (Google DeepMind), Pedro Fonseca (Purdue University), Petros Maniatis (Google DeepMind)
Paper One Simple API Can Cause Hundreds of Bugs An Analysis of Refcounting Bugs in All Modern Linux Kernels
Liang He (Institute of Software, CAS China), Purui Su (Institute of Software, CAS China), Chao Zhang (Tsinghua University), Yan Cai (Institute of Software, Chinese Academy of Sciences), Jinxin Ma (CNITSEC)
Coffee Break
Session 2: Reliability (Chair: Ding Yuan) 11:00 - 12:30 (UTC+02:00)
Paper Validating JIT Compilers via Compilation Space Exploration
Best Paper Award
Cong Li (Nanjing University), Yanyan Jiang (Nanjing University), Chang Xu (Nanjing University), Zhendong Su (ETH Zurich)
Paper Automated Verification of an In-Production DNS Authoritative Engine
Naiqian Zheng (Peking University), Mengqi Liu (Alibaba Cloud), Yuxing Xiang (Peking University), Linjian Song (Alibaba Cloud), Dong Li (Alibaba Cloud), Feng Han (Alibaba Cloud), Nan Wang (Alibaba Cloud), Yong Ma (Alibaba Cloud), Zhuo Liang (Alibaba Cloud), Dennis Cai (Alibaba Cloud), Ennan Zhai (Alibaba Cloud), Xuanzhe Liu (Peking University), Xin Jin (Peking University)
Paper Acto: Automatic End-to-End Testing for Operation Correctness of Cloud System Management
Jiawei Tyler Gu (University of Illinois at Urbana-Champaign), Xudong Sun (University of Illinois at Urbana-Champaign), Wentao Zhang (University of Illinois at Urbana-Champaign), Yuxuan Jiang (University of Illinois at Urbana-Champaign), Chen Wang (IBM Research), Mandana Vaziri (IBM Research), Owolabi Legunsen (Cornell University), Tianyin Xu (University of Illinois at Urbana-Champaign)
Paper Grove: a Separation-Logic Library for Verifying Distributed Systems
Upamanyu Sharma (Massachusetts Institute of Technology), Ralf Jung (ETH Zurich), Joseph Tassarotti (New York University), Frans Kaashoek (MIT), Nickolai Zeldovich (MIT)
Lunch 12:30 - 14:00 (UTC+02:00)
Session 3: Storage (Chair: Peter Macko) 14:00 - 15:30 (UTC+02:00)
Paper FIFO queues are all you need for cache eviction
Juncheng Yang (Carnegie Mellon University), Yazhuo Zhang (Emory University), Ziyue Qiu (Carnegie Mellon University), Yao Yue (Pelikan Foundation), Rashmi Vinayak (Carnegie Mellon University)
Paper Enabling High-Performance and Secure Userspace NVM File Systems with the Trio Architecture
Best Paper Award
Diyu Zhou (EPFL), Vojtech Aschenbrenner (EPFL), Tao Lyu (EPFL), Jian Zhang (Rutgers University), Sudarsun Kannan (Rutgers University), Sanidhya Kashyap (EPFL)
Paper Project Silica: Towards Sustainable Cloud Archival Storage in Glass
Patrick Anderson (Microsoft), Erika Blancada Aranas (Microsoft), Youssef Assaf (Microsoft), Raphael Behrendt (Microsoft), Richard Black (Microsoft), Marco Caballero (Microsoft), Pashmina Cameron (Microsoft), Burcu Canakci (Microsoft), Thales de Carvalho (Microsoft), Andromachi Chatzieleftheriou (Microsoft), Rebekah Storan Clarke (Microsoft), James Clegg (Microsoft), Daniel Cletheroe (Microsoft), Bridgette Cooper (Microsoft), Tim Deegan (Microsoft), Austin Donnelly (Microsoft), Rokas Drevinskas (Microsoft), Alexander Gaunt (Microsoft), Christos Gkantsidis (Microsoft), Ariel Gomez Diaz (Microsoft), Istvan Haller (Microsoft), Freddie Hong (Microsoft), Teodora Ilieva (Microsoft), Shashidhar Joshi (Microsoft), Russell Joyce (Microsoft), Mint Kunkel (Microsoft), David Lara (Microsoft), Sergey Legtchenko (Microsoft), Fanglin Linda Liu (Microsoft), Bruno Magalhaes (Microsoft), Alana Marzoev (Microsoft), Marvin McNett (Microsoft), Jayashree Mohan (Microsoft), Michael Myrah (Microsoft), Trong Nguyen (Microsoft), Sebastian Nowozin (Microsoft), Aaron Ogus (Microsoft), Hiske Overweg (Microsoft), Antony Rowstron (Microsoft), Maneesh Sah (Microsoft), Masaaki Sakakura (Microsoft), Peter Scholtz (Microsoft), Nina Schreiner (Microsoft), Omer Sella (Microsoft), Adam Smith (Microsoft), Ioan Stefanovici (Microsoft), David Sweeney (Microsoft), Benn Thomsen (Microsoft), Govert Verkes (Microsoft), Phil Wainman (Microsoft), Jonathan Westcott (Microsoft), Luke Weston (Microsoft), Charles Whittaker (Microsoft), Pablo Wilke Berenguer (Microsoft), Hugh Williams (Microsoft), Thomas Winkler (Microsoft), Stefan Winzeck (Microsoft)
Paper RackBlox: A Software-Defined Rack-Scale Storage System with Network-Storage Co-Design
Benjamin Reidys (University of Illinois Urbana Champaign), Yuqi Xue (University of Illinois Urbana Champaign), Daixuan Li (University of Illinois Urbana Champaign), Bharat Sukhwani (IBM Research), Wen-mei Hwu (University of Illinois Urbana Champaign), Deming Chen (University of Illinois Urbana Champaign), Sameh Asaad (IBM Research), Jian Huang (University of Illinois Urbana Champaign)
Coffee Break
Session 4: Cloud (Chair: Edouard Bugnion) 16:00 - 17:30 (UTC+02:00)
Paper Cornflakes: Zero-Copy Serialization for Microsecond-Scale Networking
Deepti Raghavan (Stanford University), Shreya Ravi (Stanford University), Gina Yuan (Stanford University), Pratiksha Thaker (Carnegie Mellon University), Sanjari Srivastava (Stanford University), Micah Murray (UC Berkeley), Pedro Henrique Penna (Microsoft Research), Amy Ousterhout (UC San Diego), Philip Levis (Stanford University and Google), Matei Zaharia (UC Berkeley), Irene Zhang (Microsoft Research)
Paper Understanding Silent Data Corruptions in a Large Production CPU Population
Shaobu Wang (Tsinghua University), Guangyan Zhang (Tsinghua University), Junyu Wei (Tsinghua University), Yang Wang (Meta/The Ohio State University), Jiesheng Wu (Alibaba Cloud), Qingchao Luo (Alibaba Cloud)
Paper XFaaS: Hyperscale and Low Cost Serverless Functions at Meta
Alireza Sahraei (Meta Platforms, Inc), Soteris Demetriou (Imperial College London, Meta Platforms), Amirali Sobhgol (Meta Platforms, Inc), Haoran Zhang (University of Pennsylvania), Abhigna Nagaraja (Meta Platforms, Inc), Neeraj Pathak (Meta Platforms, Inc), Girish Joshi (Meta Platforms, Inc), Carla Souza (Meta Platforms, Inc), Bo Huang (Meta Platforms, Inc), Wyatt Cook (Meta Platforms, Inc), Andrii Golovei (Meta Platforms, Inc), Pradeep Venkat (Meta Platforms, Inc), Andrew McFague (Meta Platforms, Inc), Dimitrios Skarlatos (Carnegie Mellon University, Meta Platforms), Vipul Patel (Meta Platforms, Inc), Ravinder Thind (Meta Platforms, Inc), Ernesto Gonzalez (Meta Platforms, Inc), Yun Jin (Meta Platforms, Inc), Chunqiang Tang (Meta Platforms, Inc)
Paper Pushing Performance Isolation Boundaries into Application with pBox
Yigong Hu (Johns Hopkins University), Gongqi Huang (Johns Hopkins University), Peng Huang (University of Michigan)
Transfer to Ehrenbreitstein Fortress
Please proceed immediately to the lower cable car station (35-minute walk [walking directions]). If you have indicated that you prefer a shuttle during registration: The shuttles will depart in front of the conference venue. 17:30 - 18:00 (UTC+02:00)
Poster Session
The poster session will happen in Fortress Ehrenbreitstein. 18:00 - 20:30 (UTC+02:00)

Wednesday, October 25, 2023

Session 5: Distributed systems (Chair: Manos Kapritsos) 08:30 - 10:00 (UTC+02:00)
Paper Flexible Advancement in Asynchronous BFT Consensus
Shengyun Liu (Shanghai Jiao Tong University), Wenbo Xu (Blockchain Platform Division, Ant Group), Chen Shan (Blockchain Platform Division, Ant Group), Xiaofeng Yan (Blockchain Platform Division, Ant Group), Tianjing Xu (Blockchain Platform Division, Ant Group), Bo Wang (Blockchain Platform Division, Ant Group), Lei Fan (Shanghai Jiao Tong University), Fuxi Deng (Blockchain Platform Division, Ant Group), Ying Yan (Blockchain Platform Division, Ant Group), Hui Zhang (Blockchain Platform Division, Ant Group)
Paper QuePaxa: Escaping the Tyranny of Timeouts in Consensus
Pasindu Tennage (Ecole Polytechnique Federale de Lausanne (EPFL)), Cristina Basescu (Ecole Polytechnique Federale de Lausanne (EPFL)), Lefteris Kokoris-Kogias (IST Austria, Mysten Labs), Ewa Syta (Trinity College), Philipp Jovanovic (UCL), Vero Estrada-Galinanes (Ecole Polytechnique Federale de Lausanne (EPFL)), Bryan Ford (Ecole Polytechnique Federale de Lausanne (EPFL))
Paper Antipode: Enforcing Cross-Service Causal Consistency in Distributed Applications
João Loff (INESC-ID, Instituto Superior Técnico, Universidade de Lisboa), Daniel Porto (INESC-ID, Instituto Superior Técnico, Universidade de Lisboa), João Garcia (INESC-ID, Instituto Superior Técnico, Universidade de Lisboa), Jonathan Mace (Max Planck Institute for Software Systems and Microsoft Research), Rodrigo Rodrigues (INESC-ID, Instituto Superior Técnico, Universidade de Lisboa)
Paper Halfmoon: Log-Optimal Fault-Tolerant Stateful Serverless Computing
Sheng Qi (Peking University), Xuanzhe Liu (Peking University), Xin Jin (Peking University)
Coffee Break
Session 6: Learning I (Chair: Timothy Roscoe) 10:30 - 12:00 (UTC+02:00)
Paper PIT: Optimization of Dynamic Sparse Deep Learning Models via Permutation Invariant Transformation
Ningxin Zheng (Microsoft Research), Huiqiang Jiang (Microsoft Research), Quanlu Zhang (Microsoft Research), Zhenhua Han (Microsoft Research), Lingxiao Ma (Microsoft Research), Yuqing Yang (Microsoft Research), Fan Yang (Microsoft Research), Chengruidong Zhang (Microsoft Research), Lili Qiu (Microsoft Research), Mao Yang (Microsoft Research), Lidong Zhou (Microsoft Research)
Paper Bagpipe: Accelerating Deep Recommendation Model Training
Saurabh Agarwal (University of Wisconsin-Madison), Chengpo Yan (University of Wisconsin-Madison), Ziyi Zhang (University of Chicago), Shivaram Venkataraman (University of Wisconsin-Madison)
Paper GEMINI: Fast Failure Recovery in Distributed Training with In-Memory Checkpoints
Zhuang Wang (Rice University), Zhen Jia (Amazon Web Services, Inc.), Shuai Zheng (Amazon Web Services), Zhen Zhang (Amazon Web Services), Xinwei Fu (Amazon Web Services), T. S. Eugene Ng (Rice University), Yida Wang (Amazon)
Paper Oobleck: Resilient Distributed Training of Large Models Using Pipeline Templates
Insu Jang (University of Michigan), Zhenning Yang (University of Michigan), Zhen Zhang (Amazon Web Services), Xin Jin (Peking University), Mosharaf Chowdhury (University of Michigan)
Lunch 12:00 - 13:30 (UTC+02:00)
Athena Lecture (Chair: Haibo Chen) 13:30 - 14:30 (UTC+02:00)
Margo Seltzer, Canada 150 Chair in Computer Systems

The Case for Collaboration (Everything a Systems Person really needs to know about Machine Learning)

It's 2023, and the answer to every system performance or optimization problem is "machine learning." But what kinds of models are appropriate for these applications? I'm going to try to convince you that, as in good system design, "simpler is better." And, in this case, simpler has many benefits: simpler models are typically more efficient in both space and time, they are frequently transparently interpretable, and they produce accuracy and generalization equivalent to the fanciest deep learning model you can build.
Session 7: Security and Privacy (Chair: Haibo Chen) 15:00 - 16:30 (UTC+02:00)
Paper Private Web Search with Tiptoe
Alexandra Henzinger (MIT), Emma Dauterman (UC Berkeley), Henry Corrigan-Gibbs (MIT), Nickolai Zeldovich (MIT)
Paper Siloz: Leveraging DRAM Isolation Domains to Prevent Inter-VM Rowhammer
Kevin Loughlin (University of Michigan), Jonah Rosenblum (University of Michigan), Stefan Saroiu (Microsoft), Alec Wolman (Microsoft), Dimitrios Skarlatos (Carnegie Mellon University), Baris Kasikci (University of Washington and Google)
Paper Edna: Disguising and Revealing User Data in Web Applications
Lillian Tsai (MIT), Hannah Gross (Brown University), Eddie Kohler (Harvard University), Frans Kaashoek (MIT), Malte Schwarzkopf (Brown University)
Paper Arboretum: A Planner for Large-Scale Federated Analytics with Differential Privacy
Elizabeth Margolin (University of Pennsylvania), Karan Newatia (University of Pennsylvania), Tao Luo (University of Pennsylvania), Edo Roth (University of Pennsylvania), Andreas Haeberlen (University of Pennsylvania and Roblox)
Coffee Break
SRC Presentations 17:00 - 18:00 (UTC+02:00)
Undergraduate Category

SRC 2. Alohomora: Practical End-to-End Privacy Compliance with Policy Containers and Privacy-Critical Regions
Artem Agvanian (Brown University)

SRC 3. Methods for Lazy Promotion on FIFO and LRU-based Eviction Algorithms
Zhuofan Chen (Carnegie Mellon University)

Graduate Category

SRC 6. Virtuoso TCP Stack: Squashing Isolation and Resource Efficiency Tradeoffs in Virtualized Environments
Matheus Stolet (MPI-SWS)

SRC 7. Full System Energy Estimation with Modular Simulation
Jonas Kaufmann (MPI-SWS)

SRC 9. Funhouse: A Hall of Mirrors Database
Hannah Gross (MIT)
Banquet Transfer
Please proceed to landing stage 5 to board the cruise ship (25-minute walk to boat dock [walking directions]). The ship will sail at 19:00 sharp. If you have indicated that you prefer a shuttle during registration: The shuttles will depart in front of the conference venue. 18:00 - 18:30 (UTC+02:00)
Banquet, Business Meeting, and Awards Presentation (On cruise ship) 19:00 - 22:00 (UTC+02:00)

Thursday, October 26, 2023

Session 8: Datacenter (Chair: Dmitrios Skarlatos) 09:00 - 10:30 (UTC+02:00)
Paper Achieving Microsecond-Scale Tail Latency Efficiently with Approximate Optimal Scheduling
Rishabh Iyer (EPFL), Musa Unal (EPFL), Marios Kogias (Imperial College London), George Candea (EPFL)
Paper Blueprint: A Toolchain for Highly-Reconfigurable Microservice Applications
Vaastav Anand (Max Planck Institute for Software Systems), Deepak Garg (Max Planck Institute for Software Systems), Antoine Kaufmann (Max Planck Institute for Software Systems), Jonathan Mace (Microsoft Research)
Paper A Cloud-Scale Characterization of Remote Procedure Calls
Korakit Seemakhupt (University of Virginia), Brent E. Stephens (Google and University of Utah), Samira Khan (Google and University of Virginia), Sihang Liu (University of Waterloo), Hassan Wassel (Google), Soheil Hassas Yeganeh (Google), Alex C. Snoeren (Google and UC San Diego), Arvind Krishnamurthy (Google and University of Washington), David Culler (Google), Henry M. Levy (Google and University of Washington)
Paper PVM: Efficient Shadow Paging for Deploying Secure Containers in Cloud-native Environment
Hang Huang (Alibaba Group), Jiangshan Lai (Ant Group), Jia Rao (The University of Texas at Arlington), Hui Lu (The University of Texas at Arlington), Wenlong Hou (Ant Group), Hang Su (Ant Group), Quan Xu (Alibaba Group), Jiang Zhong (Alibaba Group), Jiahao Zeng (Alibaba Group), Xu Wang (Ant Group), Zhengyu He (Ant Group), Weidong Han (Alibaba Group), Jiang Liu (Alibaba Group), Tao Ma (Alibaba Group), Song Wu (Huazhong University of Science and Technology)
Coffee Break
Session 9: Data and databases (Chair: Doug Terry) 11:00 - 12:30 (UTC+02:00)
Paper Falcon: Fast OLTP Engine for Persistent Cache and Non-Volatile Memory
Zhicheng Ji (Tsinghua University), Kang Chen (Tsinghua University and Zhongguancun Laboratory), Leping Wang (Tsinghua University), Mingxing Zhang (Tsinghua University), Yongwei Wu (Tsinghua University)
Paper SPFresh: Incremental In-Place Update for Billion-Scale Vector Search
Yuming Xu (University of Science and Technology of China & Microsoft Research), Hengyu Liang (University of Science and Technology of China), Jin Li (Harvard University), Shuotao Xu (Microsoft Research), Qi Chen (Microsoft Research), Qianxi Zhang (Microsoft Research), Cheng Li (University of Science and Technology of China), Ziyue Yang (Microsoft Research), Fan Yang (Microsoft Research Asia), Yuqing Yang (Microsoft Research), Peng Cheng (Microsoft Research), Mao Yang (Microsoft Research)
Paper gSampler: General and Efficient GPU-based Graph Sampling for Graph Learning
Ping Gong (University of Science and Technology of China), Renjie Liu (Southern University of Science and Technology), Zunyao Mao (Southern University of Science and Technology), Zhenkun Cai (AWS AI Shanghai Lab), Xiao Yan (Southern University of Science and Technology), Cheng Li (University of Science and Technology of China), Minjie Wang (AWS AI Shanghai Lab), Zhuozhao Li (Southern University of Science and Technology)
Paper Turbo: Effective Caching in Differentially-Private Databases
Kelly Kostopoulou (Columbia University), Pierre Tholoniat (Columbia University), Asaf Cidon (Columbia University), Roxana Geambasu (Columbia University), Mathias Lécuyer (University of British Columbia)
Lunch 12:30 - 14:00 (UTC+02:00)
Session 10: Learning II (Chair: Andrew Quinn) 14:00 - 15:30 (UTC+02:00)
Paper Paella: Low-latency Model Serving with Software-defined GPU Scheduling
Kelvin K.W. Ng (University of Pennsylvania), Henri Maxime Demoulin (DBOS, inc), Vincent Liu (University of Pennsylvania)
Paper Efficient Memory Management for Large Language Model Serving with PagedAttention
Woosuk Kwon (UC Berkeley), Zhuohan Li (UC Berkeley), Siyuan Zhuang (UC Berkeley), Ying Sheng (Stanford University), Lianmin Zheng (UC Berkeley), Cody Hao Yu (Independent Researcher), Joseph Gonzalez (UC Berkeley), Hao Zhang (UC San Diego), Ion Stoica (UC Berkeley)
Paper UGACHE: A Unified GPU Cache for Embedding-based Deep Learning
Xiaoniu Song (Shanghai Jiao Tong University; Shanghai Artificial Intelligence Laboratory), Yiwen Zhang (Shanghai Jiao Tong University), Rong Chen (Shanghai Jiao Tong University; Shanghai Artificial Intelligence Laboratory), Haibo Chen (Shanghai Jiao Tong University)
Paper Sia: Heterogeneity-aware, goodput-optimized ML-cluster scheduling
Suhas Jayaram Subramanya (Carnegie Mellon University), Daiyaan Arfeen (Carnegie Mellon University), Shouxu Lin (Cornell University), Aurick Qiao (Petuum, Inc.), Zhihao Jia (Carnegie Mellon University), Gregory R. Ganger (Carnegie Mellon University)
Coffee Break
Session 11: Distributed and disaggregated memory (Chair: Baris Kasikci) 16:00 - 17:30 (UTC+02:00)
Paper Partial Failure Resilient Memory Management System for (CXL-based) Distributed Shared Memory
Mingxing Zhang (Tsinghua University), Teng Ma (Alibaba Group), Jinqi Hua (Tsinghua University), Zheng Liu (Alibaba Group), Kang Chen (Tsinghua University), Ning Ding (Alibaba Group), Fan Du (Intel), Jinlei Jiang (Tsinghua University), Tao Ma (Alibaba Group), Yongwei Wu (Tsinghua University)
Paper Ditto: An Elastic and Adaptive Memory-Disaggregated Caching System
Jiacheng Shen (The Chinese University of Hong Kong), Pengfei Zuo (Huawei Cloud), Xuchuan Luo (Fudan University), Yuxin Su (Sun Yat-sen University), Jiazhen Gu (The Chinese University of Hong Kong), Hao Feng (Huawei Cloud), Yangfan Zhou (Fudan University), Michael R. Lyu (The Chinese University of Hong Kong)
Paper Mira: A Program-Behavior-Guided Far Memory System
Zhiyuan Guo (University of California, San Diego), Zijian He (University of California, San Diego), Yiying Zhang (University of California, San Diego)
Closing 17:30 - 17:40 (UTC+02:00)
Sponsor Event: Cocktail party organized by Huawei; all conference attendees are invited
Please register here. 18:30 - 22:00 (UTC+02:00)