Table of Contents

• LithOS: An Operating System for Efficient Machine Learning on GPUsPatrick H. Coppock, Brian Zhang, Eliot H. Solomon, Vasilis Kypriotis, Leon Yang, Bikash Sharma, Dan Schatzberg, Todd C. Mowry, Dimitrios Skarlatos 0002. 1-17 [doi]
• μFork: Supporting POSIX fork Within a Single-Address-Space OSJohn Alistair Kressel, Hugo Lefeuvre, Pierre Olivier. 18-35 [doi]
• Tock: From Research To Securing 10 Million ComputersLeon Schuermann, Brad Campbell, Branden Ghena, Philip Alexander Levis, Amit Levy, Pat Pannuto. 36-49 [doi]
• Proto: A Guided Journey through Modern OS ConstructionWonkyo Choe, Rongxiang Wang, Afsara Benazir, Felix Xiaozhu Lin. 50-66 [doi]
• CHERIoT RTOS: An OS for Fine-Grained Memory-Safe Compartments on Low-Cost Embedded DevicesSaar Amar, Tony Chen, David Chisnall, Nathaniel Wesley Filardo, Ben Laurie, Hugo Lefeuvre, Kunyan Liu, Simon W. Moore, Robert Norton-Wright, Margo I. Seltzer, Yucong Tao, Robert N. M. Watson, Hongyan Xia. 67-84 [doi]
• The Design and Implementation of a Virtual Firmware MonitorCharly Castes, François Costa, Neelu S. Kalani, Timothy Roscoe, Nate Foster, Thomas Bourgeat, Edouard Bugnion. 85-100 [doi]
• Oasis: Pooling PCIe Devices Over CXL to Boost UtilizationYuhong Zhong, Daniel S. Berger, Pantea Zardoshti, Enrique Saurez, Jacob Nelson 0001, Dan R. K. Ports, Antonis Psistakis, Joshua Fried, Asaf Cidon. 101-119 [doi]
• Spirit: Fair Allocation of Interdependent Resources in Remote Memory SystemsSeungSeob Lee, Jachym Putta, Ziming Mao, Anurag Khandelwal. 120-135 [doi]
• Scalable Far Memory: Balancing Faults and EvictionsYueyang Pan, Yash Lala, Musa Unal, Yujie Ren, SeungSeob Lee, Abhishek Bhattacharjee, Anurag Khandelwal, Sanidhya Kashyap. 136-152 [doi]
• Device-Assisted Live Migration of RDMA DevicesArtem Y. Polyakov, Gal Shalom, Asaf Schwartz, Aviad Yehezkel, Omri Ben David, Omri Kahalon, Ariel Shahar, Liran Liss. 153-168 [doi]
• Demeter: A Scalable and Elastic Tiered Memory Solution for Virtualized Cloud via Guest DelegationJunliang Hu, Zhisheng Hu, Chun-Feng Wu, Ming-Chang Yang. 169-185 [doi]
• Robust LLM Training Infrastructure at ByteDanceBorui Wan, Gaohong Liu, Zuquan Song, Jun Wang, Yun Zhang, Guangming Sheng, Shuguang Wang, Houmin Wei, Chenyuan Wang, Weiqiang Lou, Xi Yang, Mofan Zhang, Kaihua Jiang, Cheng Ren, Xiaoyun Zhi, Menghan Yu, Zhe Nan, Zhuolin Zheng, Baoquan Zhong, Qinlong Wang, Huan Yu, Jinxin Chi, Wang Zhang, Yuhan Li, Zixian Du, Sida Zhao, Yongqiang Zhang, Jingzhe Tang, Zherui Liu, Chuan Wu, Yanghua Peng, Haibin Lin, Wencong Xiao, Xin Liu, Liang Xiang. 186-203 [doi]
• Sailor: Automating Distributed Training over Dynamic, Heterogeneous, and Geo-distributed ClustersFoteini Strati, Zhendong Zhang, George Manos, Ixeia Sánchez Périz, Qinghao Hu 0004, Tiancheng Chen, Berk Buzcu, Song Han 0003, Pamela Delgado, Ana Klimovic. 204-220 [doi]
• DCP: Addressing Input Dynamism In Long-Context Training via Dynamic Context ParallelismChenyu Jiang 0002, Zhenkun Cai, Ye Tian, Zhen Jia 0001, Yida Wang 0003, Chuan Wu 0001. 221-236 [doi]
• TrainVerify: Equivalence-Based Verification for Distributed LLM TrainingYunchi Lu, Youshan Miao, Cheng Tan 0005, Peng Huang 0005, Yi Zhu, Xian Zhang 0001, Fan Yang 0024. 237-253 [doi]
• Mycroft: Tracing Dependencies in Collective Communication Towards Reliable LLM TrainingYangtao Deng, Lei Zhang, Qinlong Wang, Xiaoyun Zhi, Xinlei Zhang, Zhuo Jiang, Haohan Xu, Lei Wang, Zuquan Song, Gaohong Liu, Yang Bai, Shuguang Wang, Wencong Xiao, Jianxi Ye, Minlan Yu, Hong Xu. 254-269 [doi]
• Mitigating Application Resource Overload with Targeted Task CancellationYigong Hu, Zeyin Zhang, Yicheng Liu, Yile Gu, Shuangyu Lei, Baris Kasikci, Peng Huang. 270-285 [doi]
• Orthrus: Efficient and Timely Detection of Silent User Data Corruption in the Cloud with Resource-Adaptive Computation ValidationChenxiao Liu, Zhenting Zhu, Quanxi Li, Yanwen Xia, Yifan Qiao, Xiangyun Deng, Youyou Lu, Tao Xie 0001, Huimin Cui, Zidong Du, Harry Xu 0001, Chenxi Wang 0005. 286-304 [doi]
• Optimistic Recovery for High-Availability Software via Partial Process State PreservationYuzhuo Jing, Yuqi Mai, Angting Cai, Yi Chen, Wanning He, Xiaoyang Qian, Peter M. Chen, Peng Huang. 305-321 [doi]
• COpter: Efficient Large-Scale Resource-Allocation via Continual OptimizationSuhas Jayaram Subramanya, Don Kurian Dennis, Virginia Smith, Gregory R. Ganger. 322-340 [doi]
• Fast End-to-End Performance Simulation of Accelerated Hardware-Software StacksJiacheng Ma 0002, Jonas Kaufmann, Emilien Guandalino, Rishabh R. Iyer, Thomas Bourgeat, George Candea. 341-358 [doi]
• Characterizing Mobile SoC for Accelerating Heterogeneous LLM InferenceLe Chen, Dahu Feng, Erhu Feng, Yingrui Wang, Rong Zhao, Yubin Xia, Pinjie Xu, Haibo Chen 0001. 359-374 [doi]
• IC-Cache: Efficient Large Language Model Serving via In-context CachingYifan Yu, Yu Gan, Nikhil Sarda, Lillian Tsai, Jiaming Shen, Yanqi Zhou, Arvind Krishnamurthy, Fan Lai 0001, Hank Levy, David E. Culler. 375-398 [doi]
• PrefillOnly: An Inference Engine for Prefill-only Workloads in Large Language Model ApplicationsKuntai Du, Bowen Wang, Chen Zhang 0001, Yiming Cheng, Qing Lan, Hejian Sang, Yihua Cheng, Jiayi Yao, Xiaoxuan Liu, Yifan Qiao, Ion Stoica, Junchen Jiang. 399-414 [doi]
• Pie: A Programmable Serving System for Emerging LLM ApplicationsIn Gim, Zhiyao Ma, SeungSeob Lee, Lin Zhong 0001. 415-430 [doi]
• DiffKV: Differentiated Memory Management for Large Language Models with Parallel KV CompactionYanqi Zhang, Yuwei Hu, Runyuan Zhao, John C. S. Lui, Haibo Chen. 431-445 [doi]
• Jenga: Effective Memory Management for Serving LLM with HeterogeneityChen Zhang 0001, Kuntai Du, Shu Liu, Woosuk Kwon, Xiangxi Mo, Yufeng Wang, Xiaoxuan Liu, Kaichao You, Zhuohan Li 0001, Mingsheng Long, Jidong Zhai, Joseph Gonzalez 0001, Ion Stoica. 446-461 [doi]
• cache_ext: Customizing the Page Cache with eBPFTal Zussman, Ioannis Zarkadas, Jeremy Carin, Andrew Cheng, Hubertus Franke, Jonas Pfefferle, Asaf Cidon. 462-478 [doi]
• Aeolia: A Fast and Secure Userspace Interrupt-Based Storage StackChuandong Li 0004, Ran Yi, Zonghao Zhang, Jing Liu, Changwoo Min, Jie Zhang, Yingwei Luo, Xiaolin Wang 0001, Zhenlin Wang, Diyu Zhou. 479-495 [doi]
• Sleeping with One Eye Open: Fast, Sustainable Storage with SandmanYanbo Zhou, Erci Xu, Anisa Su, Jim Harris, Adam Manzanares, Steven Swanson. 496-511 [doi]
• Loom: Efficient Capture and Querying of High-Frequency TelemetryFranco Solleza, Shihang Li, William Sun, Richard Tang, Malte Schwarzkopf, Andrew Crotty, David Cohen, Nesime Tatbul, Stan Zdonik. 512-528 [doi]
• Pesto: Cooking up High Performance BFT QueriesFlorian Suri-Payer, Neil Giridharan, Liam Arzola, Shir Cohen, Lorenzo Alvisi, Natacha Crooks. 529-554 [doi]
• Tiga: Accelerating Geo-Distributed Transactions with Synchronized ClocksJinkun Geng, Shuai Mu, Anirudh Sivaraman, Balaji Prabhakar. 555-571 [doi]
• Tempo: Compiled Dynamic Deep Learning with Symbolic Dependence GraphsPedro F. Silvestre, Peter R. Pietzuch. 572-588 [doi]
• SAND: A New Programming Abstraction for Video-based Deep LearningJuncheol Ye, Seungkook Lee, Hwijoon Lim, Jihyuk Lee, Uitaek Hong, Youngjin Kwon, Dongsu Han. 589-605 [doi]
• METIS: Fast Quality-Aware RAG Systems with Configuration AdaptationSiddhant Ray, Rui Pan 0003, Zhuohan Gu, Kuntai Du, Shaoting Feng, Ganesh Ananthanarayanan, Ravi Netravali, Junchen Jiang. 606-622 [doi]
• HedraRAG: Co-Optimizing Generation and Retrieval for Heterogeneous RAG WorkflowsZhengding Hu, Vibha Murthy, Zaifeng Pan, Wanlu Li, Xiaoyi Fang, Yufei Ding 0001, Yuke Wang. 623-638 [doi]
• Coyote v2: Raising the Level of Abstraction for Data Center FPGAsBenjamin Ramhorst, Dario Korolija, Maximilian Jakob Heer, Jonas Dann, Luhao Liu, Gustavo Alonso. 639-654 [doi]
• KNighter: Transforming Static Analysis with LLM-Synthesized CheckersChenyuan Yang, Zijie Zhao, Zichen Xie, Haoyu Li, Lingming Zhang. 655-669 [doi]
• Fawkes: Finding Data Durability Bugs in DBMSs via Recovered Data State VerificationZhiyong Wu 0010, Jie Liang 0006, Jingzhou Fu, Wenqian Deng, Yu Jiang 0001. 670-684 [doi]
• Ghost in the Android Shell: Pragmatic Test-oracle Specification of a Production HypervisorKayvan Memarian, Ben Simner, David Kaloper-Mersinjak, Thibaut Pérami, Peter Sewell. 685-700 [doi]
• eBPF Misbehavior Detection: Fuzzing with a Specification-Based OracleTao Lyu 0004, Kumar Kartikeya Dwivedi, Thomas Bourgeat, Mathias Payer, Meng Xu 0001, Sanidhya Kashyap. 701-718 [doi]
• WASIT: Deep and Continuous Differential Testing of WebAssembly System Interface ImplementationsYage Hu, Wen Zhang, Botang Xiao, Qingchen Kong, Boyang Yi, Suxin Ji, Songlan Wang, Wenwen Wang 0001. 719-735 [doi]
• Prove It to the Kernel: Precise Extension Analysis via Proof-Guided Abstraction RefinementHao Sun 0021, Zhendong Su 0001. 736-751 [doi]
• Atmosphere: Practical Verified Kernels with Rust and VerusXiangdong Chen, Zhaofeng Li 0004, Jerry Zhang, Vikram Narayanan, Anton Burtsev. 752-767 [doi]
• AutoMan: Facilitating Verified Distributed Systems Development Through Automatic Code Generation and Manual OptimizationsZihao Zhang, Ti Zhou, Christa Jenkins, Omar Chowdhury, Shuai Mu 0001. 768-785 [doi]
• TickTock: Verified Isolation in a Production Embedded OSVivien Rindisbacher, Evan Johnson 0001, Nico Lehmann, Tyler Potyondy, Pat Pannuto, Stefan Savage, Deian Stefan, Ranjit Jhala. 786-801 [doi]
• ORQ: Complex Analytics on Private Data with Strong Security GuaranteesEli Baum, Sam Buxbaum, Nitin Mathai, Muhammad Faisal 0001, Vasiliki Kalavri, Mayank Varia, John Liagouris. 802-833 [doi]
• TRIP: Coercion-resistant Registration for E-Voting with Verifiability and Usability in VotegralLouis-Henri Merino, Simone Colombo 0002, Rene Reyes, Alaleh Azhir, Shailesh Mishra, Pasindu Tennage, Mohammad Amin Raeisi, Haoqian Zhang, Jeff R. Allen, Bernhard Tellenbach, Vero Estrada-Galiñanes, Bryan Ford. 834-874 [doi]
• Moirai: Optimizing Placement of Data and Compute in Hybrid CloudsZiyue Qiu, Hojin Park, Jing Zhao, Yu-Kai Wang, Arnav Balyan, Gurmeet Singh, Yangjun Zhang, Suqiang (Jack) Song, Gregory R. Ganger, George Amvrosiadis. 875-891 [doi]
• Tai Chi: A General High-Efficiency Scheduling Framework for SmartNICs in Hyperscale CloudsBang Di, Yun Xu, Kaijie Guo, Yibin Shen, Yu Li, Sanchuan Cheng, Hao Zheng, Fudong Qiu, Xiaokang Hu, Naixuan Guan, Dongdong Huang, Jinhu Li, Yi Wang, Yifang Yang, Jintao Li, Hang Yang, Chen Liang, Yilong Lv, Zikang Chen, Zhenwei Lu, Xiaohan Ma, Jiesheng Wu. 892-906 [doi]
• Quilt: Resource-aware Merging of Serverless WorkflowsYuxuan Zhang, Sebastian Angel. 907-927 [doi]
• Mantle: Efficient Hierarchical Metadata Management for Cloud Object Storage ServicesJiahao Li, Biao Cao, Jielong Jian, Cheng Li 0001, Sen Han, Yiduo Wang 0002, Yufei Wu 0011, Kang Chen, Zhihui Yin, Qiushi Chen, Jiwei Xiong, Jie Zhao, Fengyuan Liu, Yan Xing, Liguo Duan, Miao Yu, Ran Zheng, Feng Wu, Xianjun Meng. 928-943 [doi]
• Unlocking True Elasticity for the Cloud-Native Era with DandelionTom Kuchler, Pinghe Li, Yazhuo Zhang, Lazar Cvetkovic, Boris Goranov, Tobias Stocker, Leon Thomm, Simone Kalbermatter, Tim Notter, Andrea Lattuada 0001, Ana Klimovic. 944-961 [doi]
• Running Consistent Applications Closer to Users with Radical for Lower LatencyNicolaas Kaashoek, Oleg A. Golev, Austin T. Li, Amit Levy, Wyatt Lloyd. 962-978 [doi]
• Managing Scalable Direct Storage Accesses for GPUs with GoFSShaobo Li 0005, Yirui Eric Zhou, Yuqi Xue, Yuan Xu, Jian Huang 0006. 979-995 [doi]
• PhoenixOS: Concurrent OS-level GPU Checkpoint and Restore with Validated SpeculationXingda Wei, Zhuobin Huang, Tianle Sun, Yingyi Hao, Rong Chen 0001, Mingcong Han, Jinyu Gu 0001, Haibo Chen 0001. 996-1013 [doi]
• KTransformers: Unleashing the Full Potential of CPU/GPU Hybrid Inference for MoE ModelsHongtao Chen, Weiyu Xie, Boxin Zhang, Jingqi Tang, Jiahao Wang, Jianwei Dong, Shaoyuan Chen, Ziwei Yuan, Chen Lin, Chengyu Qiu, Yuening Zhu, Qingliang Ou, Jiaqi Liao, Xianglin Chen, Zhiyuan Ai, Yongwei Wu 0001, Mingxing Zhang. 1014-1029 [doi]
• Aegaeon: Effective GPU Pooling for Concurrent LLM Serving on the MarketYuxing Xiang, Xue Li 0024, Kun Qian, Yufan Yang, Diwen Zhu, Wenyuan Yu, Ennan Zhai, Xuanzhe Liu, Xin Jin 0008, Jingren Zhou 0001. 1030-1045 [doi]
• Mercury: Unlocking Multi-GPU Operator Optimization for LLMs via Remote Memory SchedulingYue Guan, Xinwei Qiang, Zaifeng Pan, Daniels Johnson, Yuanwei Fang, Keren Zhou 0001, Yuke Wang, Wanlu Li, Yufei Ding 0001, Adnan Aziz. 1046-1061 [doi]
• How to Copy Memory? Coordinated Asynchronous Copy as a First-Class OS ServiceJingkai He, Yunpeng Dong, Dong Du 0003, Mo Zou, Zhitai Yu, Yuxin Ren 0001, Ning Jia, Yubin Xia, Haibo Chen 0001. 1062-1081 [doi]
• CortenMM: Efficient Memory Management with Strong Correctness GuaranteesJunyang Zhang, Xiangcan Xu, Yonghao Zou, Zhe Tang, Xinyi Wan, Kang Hu, Siyuan Wang, Wenbo Xu, Di Wang, Hao Chen 0023, Lin Huang, Shoumeng Yan, Yuval Tamir, Yingwei Luo, Xiaolin Wang, Huashan Yu, Zhenlin Wang, Hongliang Tian, Diyu Zhou. 1082-1098 [doi]
• Rearchitecting the Thread Model of In-Memory Key-Value Stores with μTPSYoumin Chen, Jiwu Shu, Yanyan Shen, Linpeng Huang, Hong Mei 0001. 1099-1114 [doi]
• FlexGuard: Fast Mutual Exclusion Independent of SubscriptionVictor Laforet, Sanidhya Kashyap, Calin Iorgulescu, Julia Lawall, Jean-Pierre Lozi. 1115-1130 [doi]
• Scalable Address Spaces using Concurrent Interval SkiplistTae-Woo Kim, Youngjin Kwon, Jeehoon Kang. 1131-1148 [doi]
• Analyzing and Enhancing ArckFS: An Anecdotal Example of Benefits of Artifact EvaluationJonguk Jeon, Subeen Park, Sanidhya Kashyap, Sudarsun Kannan, Diyu Zhou, Jeehoon Kang. 1149-1157 [doi]