NSDI '25
as the organisers put it
The 22nd USENIX Symposium on Networked Systems Design and Implementation (NSDI '25) will take place April 28–30, 2025, at the Philadelphia Marriott Downtown in Philadelphia, PA, USA. NSDI focuses on the design principles, implementation, and practical evaluation of networked and distributed systems.
who it's for
distributed systems engineers, network engineers, cloud infrastructure engineers, machine learning infrastructure engineers, data center operators, storage engineers, smartNIC developers, security and privacy engineers
topics
programmable switch telemetry routing; smartNIC multi-target compiler; software control for RDMA; serverless pay-for-use billing; learning-based congestion control; resilient AllReduce for deep learning; gpu-disaggregated model serving; transport for in-network computing; service mesh end-to-end testing; edge video analytics enhancement; federated analytics engineering; high-throughput DAG BFT consensus
| title | author | length |
|---|---|---|
| PRED: Performance-oriented Random Early Detection for Consistently Stable Performance... | Xinle Du, Tong Li, Guangmeng Zhou, Zhuotao Liu, Hanlin Huang, Xiangyu Gao, Mowei Wang, Kun Tan, Ke Xu | 16m |
PRED implements on- switch flow counting and two loosely coupled subsystems, Flow Concurrent Stabilizer (FCS) and Queue Length Adjuster (QLA), using MIMD to… | ||
| Rajomon: Decentralized and Coordinated Overload Control for Latency-Sensitive... | Jiali Xing, Akis Giannoukos, Paul Loh, Shuyue Wang, Justin Qiu, Henri Maxime Demoulin, Konstantinos Kallas, Benjamin C. Lee | 16m |
RAJOMON is a market-style, decentralized overload controller where clients attach tokens and prices to requests and services charge per-interface prices… | ||
| Learnings from Deploying Network QoS Alignment to Application Priorities for Storage... | Matthew Buckley, Parsa Pazhooheshy, Z. Morley Mao, Nandita Dukkipati, Hamid Hajabdolali Bazzaz, Priyaranjan Jha, Yingjie Bi, Steve Middlekauff, Yashar Ganjali | 16m |
Deploys an RPC-level mapping from application priorities to DSCP/WFQ classes (Aquidus) across storage traffic, performed via per-customer, per-region canaries… | ||
| DISC: Backpressure Mitigation In Multi-tier Applications With Distributed Shared... | Brice Ekane, Djob Mvondo, Renaud Lachaize, Yérom-David Bromberg, Alain Tchana, Daniel Hagimont | 15m |
DISC lets multiple tiers jointly act as endpoints of the shared TCP connection so backend servers can emit large response bodies directly to clients while… | ||
| Enabling Silent Telemetry Data Transmission with InvisiFlow | Yinda Zhang, Liangcheng Yu, Gianni Antichi, Ran Ben Basat, Vincent Liu | 16m |
InvisiFlow routes low-priority telemetry off the shortest path by probing neighbors with pull requests and forwarding along a congestion gradient computed from… | ||
| Unlocking ECMP Programmability for Precise Traffic Control | Yadong Liu, Yunming Xiao, Xuan Zhang, Weizhen Dang, Huihui Liu, Xiang Li, Zekun He, Jilong Wang, Aleksandar Kuzmanovic, Ang Chen, Congcong Miao | 17m |
P-ECMP implements a programming model, compiler, and runtime that reuses seldom-used ECMP groups to deterministically steer flows; compiler inputs are topology… | ||
| Enabling Portable and High-Performance SmartNIC Programs with Alkali | Jiaxin Lin, Zhiyuan Guo, Mihir Shah, Tao Ji, Yiying Zhang, Daehyeok Kim, Aditya Akella | 16m |
Alkali is a multi-target SmartNIC compiler that turns a single-threaded C program into an alphair IR handler graph and uses an iterative two-stage optimizer—a… | ||
| Scaling IP Lookup to Large Databases using the CRAM Lens | Robert Chang, Pradeep Dogga, Andy Fingerhut, Victor Rios, George Varghese | 16m |
A formal CRAM lens evaluation shows concrete mappings and tradeoffs for RMT/DRMT chips: a DAG-based CRA model plus eight optimization idioms predicts memory… | ||
| Quicksand: Harnessing Stranded Datacenter Resources with Granular Computing | Zhenyuan Ruan, Shihang Li, Kaiyan Fan, Seo Jin Park, Marcos K. Aguilera, Adam Belay, Malte Schwarzkopf | 17m |
Quicksand decomposes applications into small resource proclets—separate compute and memory units—and automatically splits, merges, and migrates them to fill… | ||
| Beehive: A Scalable Disaggregated Memory Runtime Exploiting Asynchrony of Multithreaded.. | Quanxi Li, Hong Huang, Ying Liu, Yanwen Xia, Jie Zhang, Mosong Zhou, Xiaobing Feng, Huimin Cui, Quan Chen, Yizhou Shan, Chenxi Wang | 13m |
Beehive builds a Rust compile-time transform plus a runtime that turns synchronous multithreaded code into asynchronous pararoutine units and a lightweight… | ||
| Making Serverless Pay-For-Use a Reality with Leopard | Tingjia Cao, Andrea C. Arpaci-Dusseau, Remzi H. Arpaci-Dusseau, Tyler Caraza-Harter | 18m |
Shows that current serverless billing overcharges by charging execution time times a static memory cap, and introduces NPFU billing model that adds CPU caps… | ||
| GRANNY: Granular Management of Compute-Intensive Applications in the Cloud | Carlos Segarra, Simon Shillaker, Guo Li, Eleftheria Mappoura, Rodrigo Bruno, Lluís Vilanova, Peter Pietzuch | 18m |
A distributed runtime called GRANNY uses a WebAssembly-based execution abstraction to run one-vCPU Granules that offer either thread (OpenMP) or process (MPI)… | ||
| On Temporal Verification of Stateful P4 Programs | Delong Zhang, Chong Ye, Fei He | 15m |
Introduces P4LTL, a packet-processing LTL for expressing safety (e.g., no implicit drops when a register count is in-range) and liveness (per-port activation… | ||
| NDD: A Decision Diagram for Network Verification | Zechun Li, Peng Zhang, Yichi Zhang, Hongkun Yang | 15m |
Introduces Network Decision Diagram (NDD), a two-layer structure that groups header bits into fields and builds per-field BDDs while assembling them with nodes… | ||
| Smart Casual Verification of the Confidential Consortium Framework | Heidi Howard, Markus A Kuppe, Edward Ashton, Amaury Chamayou, Natacha Crooks | 18m |
Specifies CCF's distributed consensus and client-consistency model in TLA+, instruments the C++ code to emit tiny JSON state traces, and repurposes the TLC… | ||
| VEP: A Two-stage Verification Toolchain for Full eBPF Programmability | Xiwei Wu, Yueyang Feng, Tianyi Huang, Xiaoyang Lu, Shengkai Lin, Lihan Xie, Shizhen Zhao, Qinxiang Cao | 17m |
VEP is a two-stage verifier: VEP-C (annotation-guided C verifier), a VEP-compiler that preserves annotations through compilation, and VEP-eBPF, a lightweight… | ||
| MeshTest: End-to-End Testing for Service Mesh Traffic Management | Naiqian Zheng, Tianshuo Qiao, Xuanzhe Liu, Xin Jin | 14m |
MeshTest automatically generates end-to-end service mesh configs using a novel Service Flow Exploration to enumerate resource interactions, then translates… | ||
| Preventing Network Bottlenecks: Accelerating Datacenter Services with Hotspot-Aware... | Hamid Hajabdolali Bazzaz, Yingjie Bi, Weiwu Pang, Minlan Yu, Ramesh Govindan, Neal Cardwell, Nandita Dukkipati, Meng-Jung Tsai, Chris DeForeest, Yuxue Jin, Charles Carver, Jan Kopanski, Liqun Cheng, Amin Vahdat | 18m |
Analysis of Google datacenter ToR hotspots shows links that persist hot for hours and often exceed 75–90% utilization, causing over 2× end-to-end latency for… | ||
| Enhancing Network Failure Mitigation with Performance-Aware Ranking | Pooria Namyar, Arvin Ghavidel, Daniel Crankshaw, Daniel S. Berger, Kevin Hsieh, Srikanth Kandula, Ramesh Govindan, Behnaz Arzani | 16m |
Swarm ranks mitigation actions by directly estimating end-to-end flow metrics using fast approximations: epoched traffic samples, splitting short/long flows… | ||
| One-Size-Fits-None: Understanding and Enhancing Slow-Fault Tolerance in Modern... | Ruiming Lu, Yunchi Lu, Yuxuan Jiang, Guangtao Xue, Peng Huang | 17m |
A systematic testing pipeline injects diverse slow faults into six modern distributed systems and sweeps ~100k fault combinations including 8000 configuration… | ||
| Pyrrha: Congestion-Root-Based Flow Control to Eliminate Head-of-Line Blocking in... | Kexin Liu, Zhaochen Zhang, Chang Liu, Yizhi Wang, Vamsi Addanki, Stefan Schmid, Qingyue Wang, Wei Chen, Xiaoliang Wang, Jiaqi Zheng, Wenhao Sun, Tao Wu, Ke Meng, Fei Chen, Weiguang Wang, Bingyang Liu, Wanchun Dou, Guihai Chen, Chen Tian | 17m |
Claims a flow-control granularity tied to the congestion-root ID that provably eliminates head‑of‑line blocking by allocating isolation queues only for flows… | ||
| eTran: Extensible Kernel Transport with eBPF | Zhongjie Chen, Qingkai Meng, ChonLam Lao, Yifan Liu, Fengyuan Ren, Minlan Yu, Yang Zhou | 15m |
eTran extends eBPF to create an extensible in-kernel transport that adds new XDP hooks and a packet Q BPF map for queuing, plus XDP egress and XDP gym to… | ||
| White-Boxing RDMA with Packet-Granular Software Control | Chenxingyu Zhao, Jaehong Min, Ming Liu, Arvind Krishnamurthy | 17m |
Claims software control of RDMA transport at packet granularity while preserving line‑rate data path performance, implemented on Nvidia BlueField-3 DPA and… | ||
| SIRD: A Sender-Informed, Receiver-Driven Datacenter Transport Protocol | Konstantinos Prasopoulos, Ryan Kosta, Edouard Bugnion, Marios Kogias | 17m |
SIRD combines receiver-driven downlink scheduling with reactive control on shared links, implemented on Caladan and evaluated in a 144‑host simulator and a… | ||
| Accelerating Design Space Exploration for LLM Training Systems with Multi-experiment... | Fei Gui, Kaihui Gao, Li Chen, Dan Li, Vincent Liu, Ran Zhang, Hongbing Yang, Dian Xiong | 16m |
Presents Multiverse, a GPU-based LLM-training simulator that runs many independent experiments in one process using SPME and data-oriented DoD modeling, with… | ||
| Optimizing RLHF Training for Large Language Models with Stage Fusion | Yinmin Zhong, Zili Zhang, Bingyang Wu, Shengyu Liu, Yukun Chen, Changyi Wan, Hanpeng Hu, Lei Xia, Ranchen Ming, Yibo Zhu, Xin Jin | 13m |
RLHFuse claims up to 3.7× end-to-end RLHF training throughput by splitting tasks into sample-level subtasks and microbatch subtasks. It overlaps generation and… | ||
| Minder: Faulty Machine Detection for Large-scale Distributed Model Training | Yangtao Deng, Xiang Shi, Zhuo Jiang, Xingjian Zhang, Lei Zhang, Zhang Zhang, Bo Li, Zuquan Song, Hang Zhu, Gaohong Liu, Fuliang Li, Shuguang Wang, Haibin Lin, Jianxi Ye, Minlan Yu | 14m |
Minder is an automated faulty-machine detector deployed for over a year on thousand‑machine distributed training jobs; it builds per‑metric parametric models… | ||
| Holmes: Localizing Irregularities in LLM Training with Mega-scale GPU Clusters | Zhiyi Yao, Pengbo Hu, Congcong Miao, Xuya Jia, Zuning Liang, Yuedong Xu, Chunzhi He, Hao Lu, Mingzhuo Chen, Xiang Li, Zekun He, Yachen Wang, Xianneng Zou, Junchen Jiang | 16m |
Holmes localizes slow or spiky iterations in multi‑thousand GPU LLM training by logging only communication operators, classifying abnormal operators with a… | ||
| SimAI: Unifying Architecture Design and Performance Tuning for Large-Scale Large... | Xizheng Wang, Qingxu Li, Yichi Xu, Gang Lu, Dan Li, Li Chen, Heyang Zhou, Linkang Zheng, Sen Zhang, Yikai Zhu, Yang Liu, Pengcheng Zhang, Kun Qian, Kunling He, Jiaqi Gao, Ennan Zhai, Dennis Cai, Binzhang Fu | 11m |
SimAI claims a unified, high‑precision LLM training simulator that matches real runs 98.1% alignment by hijacking training frameworks (Megatron, DeepSpeed)… | ||
| ByteCheckpoint: A Unified Checkpointing System for Large Foundation Model Development | Borui Wan, Mingji Han, Yiyao Sheng, Yanghua Peng, Haibin Lin, Mofan Zhang, Zhichao Lai, Menghan Yu, Junda Zhang, Zuquan Song, Xin Liu, Chuan Wu | 13m |
ByteCheckpoint implements a parallelism-agnostic checkpoint representation and load-time resharding so checkpoints saved on one GPU layout can be read into… | ||
| Mowgli: Passively Learned Rate Control for Real-Time Video | Neil Agarwal, Rui Pan, Francis Y. Yan, Ravi Netravali | 16m |
Mowgli trains a rate controller entirely offline from production GCC telemetry, using soft actor critic plus conservative Q-learning and distributional critics… | ||
| Dissecting and Streamlining the Interactive Loop of Mobile Cloud Gaming | Yang Li, Jiaxing Qiu, Hongyi Wang, Zhenhua Li, Feng Qian, Jing Yang, Hao Lin, Yunhao Liu, Bo Xiao, Xiaokang Qin, Tianyin Xu | 17m |
A 20,000‑record, 2,000‑fps camera measurement across eight commercial platforms reveals mobile cloud gaming interactive latency of 112–403 ms and shows network… | ||
| Region-based Content Enhancement for Efficient Video Analytics at the Edge | Weijun Wang, Liang Mi, Shaowei Cen, Haipeng Dai, Yuanchun Li, Xiaoming Fu, Yunxin Liu | 14m |
RegenHance presents a concrete system that enhances only important frame regions to cut compute and bandwidth: a macroblock-based region importance predictor… | ||
| Tooth: Toward Optimal Balance of Video QoE and Redundancy Cost by Fine-Grained FEC in... | Congkai An, Huanhuan Zhang, Shibo Wang, Jingyang Kang, Anfu Zhou, Liang Liu, Huadong Ma, Zili Meng, Delei Ma, Yusheng Dong, Xiaogang Lei | 17m |
Paper analyzes over 60,000 cloud gaming sessions and finds coarse-grained FEC both over-protects large frames (wasting bandwidth) and under-protects small… | ||
| AsTree: An Audio Subscription Architecture Enabling Massive-Scale Multi-Party... | Tong Meng, Wenfeng Li, Chao Yuan, Changqing Yan, Le Zhang | 15m |
AsTree replaces full “subscribe-to-all” audio with a cascading tree topology over SFUs and media-plane audio selection that forwards only a few loudest streams… | ||
| AutoCCL: Automated Collective Communication Tuning for Accelerating Distributed and... | Guanbin Xu, Zhihao Le, Yinhe Chen, Zhiqi Lin, Zewen Jin, Youshan Miao, Cheng Li | 15m |
AutoCCL automates low-level tuning of NCCL to cut communication overhead: it prunes NCCL's 158 knobs to 28 parameters, partitions the huge search space with a… | ||
| OptiReduce: Resilient and Tail-Optimal AllReduce for Distributed Deep Learning in the... | Ertza Warraich, Omer Shabtai, Khalid Manaa, Shay Vargaftik, Yonatan Piasetzky, Matty Kadosh, Lalith Suresh, Muhammad Shahbaz | 16m |
OptiReduce is a time‑bounded AllReduce system that trades guaranteed delivery for bounded loss and faster completion: it implements an unreliable bounded… | ||
| Efficient Direct-Connect Topologies for Collective Communications | Liangyu Zhao, Siddharth Pal, Tapan Chugh, Weiyang Wang, Jason Fantl, Prithwish Basu, Joud Khoury, Arvind Krishnamurthy | 14m |
Presents a concrete pipeline that synthesizes direct‑connect topologies and schedules, starting from small optimal bases and using line graph expansion and… | ||
| SuperServe: Fine-Grained Inference Serving for Unpredictable Workloads | Alind Khare, Dhruv Garg, Sukrit Kalra, Snigdha Grandhi, Ion Stoica, Alexey Tumanov | 16m |
SuperServe implements SubNetAct and SlackFit to serve many weight‑shared sub-net models from a single pre-trained super-network by inserting control‑flow… | ||
| Pineapple: Unifying Multi-Paxos and Atomic Shared Registers | Tigran Bantikyan, Jonathan Zarnstorff, Te-Yen Chou, Lewis Tseng, Roberto Palmieri | 19m |
Pineapple implements a unified Multi-Paxos plus atomic-register design that uses logical timestamps (tag and slot) so any replica can serve reads and blind… | ||
| Ladder: A Convergence-based Structured DAG Blockchain for High Throughput and Low Latency | Dengcheng Hu, Jianrong Wang, Xiulong Liu, Hao Xu, Xujing Wu, Muhammad Shahzad, Guyue Liu, Keqiu Li | 17m |
A structured twin-chain DAG named Ladder imposes regular convergence points so an upper chain runs parallel PoW block production while a designated convergence… | ||
| Vegeta: Enabling Parallel Smart Contract Execution in Leaderless Blockchains | Tianjing Xu, Yongqi Zhong, Yiming Zhang, Ruofan Xiong, Jingjing Zhang, Guangtao Xue, Shengyun Liu | 18m |
The paper proposes speculate-order-replay: nodes pre-execute transactions, build a dependency graph, then deterministically replay and validate read/write… | ||
| Shoal++: High Throughput DAG BFT Can Be Fast and Robust! | Balaji Arun, Zekun Li, Florian Suri-Payer, Sourav Das, Alexander Spiegelman | 16m |
Shoal++ is a DAG BFT system that cuts average commit latency from 10.5 to 4.5 message delays while keeping DAG throughput, by running three staggered DAG… | ||
| Learning Production-Optimized Congestion Control Selection for Alibaba Cloud CDN | Xuan Zeng, Haoran Xu, Chen Chen, Xumiao Zhang, Xiaoxi Zhang, Xu Chen, Guihai Chen, Yubing Qiu, Yiping Zhang, Chong Hao, Ennan Zhai | 19m |
AliCCS is a production system that learns to pick among existing congestion controllers using a GAN-based invariant representation to remove /24 hidden-state… | ||
| GPU-Disaggregated Serving for Deep Learning Recommendation Models at Scale | Lingyun Yang, Yongchen Wang, Yinghao Yu, Qizhen Weng, Jianbo Dong, Kan Liu, Chi Zhang, Yanyi Zi, Hao Li, Zechao Zhang, Nan Wang, Yu Dong, Menglei Zheng, Lanlan Xi, Xiaowei Lu, Liang Ye, Guodong Yang, Binzhang Fu, Tao Lan, Liping Zhang, Lin Qu, Wei Wang | 18m |
Prism disaggregates CPU and GPU pools via graph-level partitioning and RDMA so embedding lookups run on CPU nodes and dense nets on GPU nodes, using a fusgraph… | ||
| Evolution of Aegis: Fault Diagnosis for AI Model Training Service in Production | Jianbo Dong, Kun Qian, Pengcheng Zhang, Zhilong Zheng, Liang Chen, Fei Feng, Yichi Xu, Yikai Zhu, Gang Lu, Xue Li, Zhihui Ren, Zhicheng Wang, Bin Luo, Peng Zhang, Yang Liu, Yanqing Chen, Yu Guan, Weicheng Wang, Chaojie Yang, Yang Zhang, Man Yuan, Hanyu Zhao, Yong Li, Zihan Zhao, Shan Li, Xianlong Zeng, Zhiping Yao, Binzhang Fu, Ennan Zhai, Wei Lin, Chao Wang, Dennis Cai | 16m |
Aegis is a production fault diagnosis system for AI model training that evolves from enhanced legacy tools to a customized collective communication library… | ||
| PAPAYA Federated Analytics Stack: Engineering Privacy, Scalability and Practicality | Harish Srinivas, Graham Cormode, Mehrdad Honarkhah, Samuel Lurye, Jonathan Hehir, Lunwen He, George Hong, Ahmed Magdy, Dzmitry Huba, Kaikai Wang, Shen Guo, Shoubhik Bhattacharya | 17m |
PAPAYA is Meta’s federated analytics stack that combines trusted execution environments with on-device, SQL-based transformations and hard-coded client… | ||
| HA/TCP: A Reliable and Scalable Framework for TCP Network Functions | Haoyu Gu, Ali José Mashtizadeh, Bernard Wong | 15m |
Framework guarantees transparent active-active replication of TCP connections by duplicating and ordering packets to a replica and delivering them only when… | ||
| High-level Programming for Application Networks | Xiangfeng Zhu, Yuyao Wang, Banruo Liu, Yongtong Wu, Nikola Bojanic, Jingrong Chen, Gilbert Louis Bernstein, Arvind Krishnamurthy, Sam Kumar, Ratul Mahajan, Danyang Zhuo | 16m |
AppNet is a framework that compiles high-level generalized match-action programs for RPC processing into platform-specific code (RPC library, colocated proxy… | ||
| State-Compute Replication: Parallelizing High-Speed Stateful Packet Processing | Qiongwen Xu, Sebastiano Miano, Xiangyu Gao, Tao Wang, Adithya Murugadass, Songyuan Zhang, Anirudh Sivaraman, Gianni Antichi, Srinivas Narayana | 16m |
Introduces state-compute replication that replicates only the state-update code and attaches compact per-packet histories so each core fast-forwards private… | ||
| MTP: Transport for In-Network Computing | Tao Ji, Rohan Vardekar, Balajee Vamanan, Brent E. Stephens, Aditya Akella | 14m |
MTP is a DPDK prototype transport that natively handles pathlet mutation, intercept, reordering and delaying by treating offloads as message-oriented pathlets… | ||
| ONCache: A Cache-Based Low-Overhead Container Overlay Network | Shengkai Lin, Shizhen Zhao, Peirui Cao, Xinchi Han, Quan Tian, Wenfeng Liu, Qi Wu, Donghai Han, Xinbing Wang | 16m |
Claims a cache of per-flow processing results (routing, tunnel headers, packet-filter decisions, with checksum fixes) eliminates repetitive overlay work and… | ||
| GREEN: Carbon-efficient Resource Scheduling for Machine Learning Clusters | Kaiqiang Xu, Decang Sun, Han Tian, Junxue Zhang, Kai Chen | 17m |
GREEN monitors per-job power with NVML readouts (plus CPU estimates) and integrates carbon intensity every 30s, uses an energy efficiency optimizer to scale… | ||
| The Benefits and Limitations of User Interrupts for Preemptive Userspace Scheduling | Linsong Guo, Danial Zuberi, Tal Garfinkel, Amy Ousterhout | 16m |
User interrupts on recent Intel CPUs lower preemption overhead versus signals and avoid the unpredictable kernel path of signals, while compiler… | ||
| Securing Public Cloud Networks with Efficient Role-based Micro-Segmentation | Sathiya Kumaran Mani, Kevin Hsieh, Santiago Segarra, Ranveer Chandra, Yajie Zhou, Srikanth Kandula | 17m |
Introduces an end-to-end system, ZTS, that infers roles from large flow logs using contrastive learning with sparse labels to reach average ARRI ≈ 0.77 on real… | ||
| Mitigating Scalability Walls of RDMA-based Container Networks | Wei Liu, Kun Qian, Zhenhua Li, Feng Qian, Tianyin Xu, Yunhao Liu, Yu Guan, Shuhong Zhu, Hongfei Xu, Lanlan Xi, Chao Qin, Ennan Zhai | 15m |
ScalaCN infers RNIC internals via combinatorial causal testing and links specific config dimensions to failures, finding that increasing flow-group query cost… | ||
| Eden: Developer-Friendly Application-Integrated Far Memory | Anil Yelam, Stewart Grant, Saarth Deshpande, Nadav Amit, Radhika Niranjan Mysore, Amy Ousterhout, Marcos K. Aguilera, Alex C. Snoeren | 12m |
Eden is a far memory system that combines hardware page guards with a few programmer hints and a user‑level runtime to get app‑integrated performance with low… | ||
| Achieving Wire-Latency Storage Systems by Exploiting Hardware ACKs | Qing Wang, Jiwu Shu, Jing Wang, Yuhao Zhang | 16m |
Juneberry implements an Ordered Queue (OQ) that uses hardware ACKs as commit signals from RDMA NICs so clients commit neoex storage requests in one round trip… | ||
| ODRP: On-Demand Remote Paging with Programmable RDMA | Zixuan Wang, Xingda Wei, Jinyu Gu, Hongrui Xie, Rong Chen, Haibo Chen | 14m |
ODRP offloads fine‑grain remote paging onto the RNIC by chaining RDMA work requests and using event‑triggered worker queues to implement allocation… | ||
| Understanding and Profiling NVMe-over-TCP Using ntprof | Yuyuan Kang, Ming Liu | 17m |
ntprof is a kernel + user tool that models NVMe‑over‑TCP as nine software switches and adds tracepoints to Linux 5.15.143 to collect lossless per‑IO timelines… | ||
| Building an Elastic Block Storage over EBOFs Using Shadow Views | Sheng Jiang, Ming Liu | 17m |
Presents a deployed software telemetry layer called shadow view that samples per-port, per-pipe, and per-SSD stats and uses a lightweight counter-based sync to… | ||
| Pushing the Limits of In-Network Caching for Key-Value Stores | Gyuyeong Kim | 17m |
OrbitCache caches whole key–value pairs by keeping packet recirculation loops in the programmable data plane instead of storing values in switch memory… | ||
| CellReplay: Towards accurate record-and-replay for cellular networks | William Sentosa, Balakrishnan Chandrasekaran, P. Brighten Godfrey, Haitham Hassanieh | 16m |
CellReplay records cellular behavior with two simultaneous workloads — light packet trains and a saturator — calibrates parameters, and replays by… | ||
| Large Network UWB Localization: Algorithms and Implementation | Nakul Garg, Irtaza Shahid, Ramanujan K Sheshadri, Karthikeyan Sundaresan, Nirupam Roy | 14m |
Locate3D is a UWB-based system that uses antenna-array joint range and angle measurements to make each peer-to-peer exchange four times more informative… | ||
| Towards Energy Efficient 5G vRAN Servers | Anuj Kalia, Nikita Lazarev, Leyang Xue, Xenofon Foukas, Bozidar Radunovic, Francis Y. Yan | 17m |
RENC adjusts CPU P‑states to use sub‑second workload variation, coupling MAC layer rate limiting with ordered frequency changes so throttling happens before… | ||
| Building Massive MIMO Baseband Processing on a Single-Node Supercomputer | Xincheng Xie, Wentao Hou, Zerui Guo, Ming Liu | 15m |
MegaStation implements a baseband pipeline as instruction streams on a single-node supercomputer with a composable GPU pool and GigaIO FabreX fabric, using an… | ||
| Efficient Multi-WAN Transport for 5G with OTTER | Mary Hogan, Gerry Wan, Yiming Qiu, Sharad Agarwal, Ryan Beckett, Rachee Singh, Paramvir Bahl | 13m |
OTTER implements a multi-WAN overlay (VMs, VPN gateways, private VNETs) across an 8-region cloud topology and a controller/orchestrator/measurement-coordinator… | ||
| Verifying maximum link loads in a changing world | Tibor Schneider, Stefano Vissicchio, Laurent Vanbever | 17m |
Velo finds maximum link loads when links fail and BGP routes change by proving worst-case per-destination happens with a single-egress worst case and by… | ||
| A Layered Formal Methods Approach to Answering Queue-related Queries | Divya Raghunathan, Maria Apostolaki, Aarti Gupta | 16m |
QuASI answers whether queue-related queries (e.g., whether a queue could exceed a threshold, burst occurrence, max buffer occupancy) are possible using only… | ||
| Runtime Protocol Refinement Checking for Distributed Protocol Implementations | Ding Ding, Zhanghan Wang, Jinyang Li, Aurojit Panda | 16m |
Ellsberg implements runtime protocol refinement checking as a local per-node checker (Go implementation) that compares observed send/receive traces to a… | ||
| CEGS: Configuration Example Generalizing Synthesizer | Jianmin Liu, Li Chen, Dan Li, Yukai Miao | 17m |
CEGS is a system that uses a GNN-based Querier to find relevant device documentation examples, a GNN-based Classifier to map example snippets to target… | ||
| Suppressing BGP Zombies with Route Status Transparency | Yosef Edery Anahory, Jie Kong, Nicholas Scaglione, Justin Furuness, Hemi Leibowitz, Amir Herzberg, Bing Wang, Yossi Gilad | 15m |
Route Status Transparency (RoST) introduces a counter per prefix-interface and a shared repository so ASes can cryptographically verify that a route was… | ||
| ValidaTor: Domain Validation over Tor | Jens Frieß, Haya Schulmann, Michael Waidner | 19m |
An HTTP-based domain-validation prototype routes CA fetches through randomized Tor exit nodes using two-hop circuits and containerized workers, achieving 12… | ||
| From Address Blocks to Authorized Prefixes: Redesigning RPKI ROV with a Hierarchical... | Zedong Ni, Yinbo Xu, Hui Zou, Yanbiao Li, Guang Cheng, Gaogang Xie | 15m |
Proposes an Authorized Prefix (AP) model and implements it as h2ROV, a hierarchical hashing validator that stores bitmap-encoded ROAs and uses two hash tables… | ||
| PreAcher: Secure and Practical Password Pre-Authentication by Content Delivery Networks | Shihan Lin, Suting Chen, Yunming Xiao, Yanqi Gu, Aleksandar Kuzmanovic, Xiaowei Yang | 14m |
A three‑party protocol offloads pre‑authentication to CDNs while keeping real passwords secret, combining OPRF + LSH so the CDN filters wrong attempts and the… | ||
| ClubHeap: A High-Speed and Scalable Priority Queue for Programmable Packet Scheduling | Zhikang Chen, Haoyu Song, Zhiyu Zhang, Yang Xu, Bin Liu | 13m |
ClubHeap implements a hardware-friendly clustered heap that achieves a CPR of one so it can issue a push+pop every cycle and effectively schedule a flow each… | ||
| Self-Clocked Round-Robin Packet Scheduling | Erfan Sharafzadeh, Raymond Matson, Jean Tourrilhes, Puneet Sharma, Soudeh Ghorbani | 16m |
Shows a new scheduler called SCRR (Self-Clocked Round-Robin) with a parameter-less algorithm that replaces DRR’s quantum with self-clocking, a Linux… | ||
| Everything Matters in Programmable Packet Scheduling | Albert Gran Alcoz, Balázs Vass, Pooria Namyar, Behnaz Arzani, Gabor Retvari, Laurent Vanbever | 15m |
Presents PACKS, a quantile-based admission and queue-mapping design that approximates both PIFO queues scheduling and admission behaviors using a… | ||
| When P4 Meets Run-to-completion Architecture | Hao Zheng, Xin Yan, Wenbo Li, Jiaqi Zheng, Xiaoliang Wang, Qingqing Zhao, Luyou He, Xiaofei Lai, Feng Gao, Fuguang Huang, Wanchun Dou, Guihai Chen, Chen Tian | 15m |
P4RTC exposes P4 on a run‑to‑completion chip (Huawei 9800 FA1C prototype, 1.2 Tbps, 8 GB HBM) and adds concrete extensions: foreground pipeline and background… | ||
| CATO: End-to-End Optimization of ML-Based Traffic Analysis Pipelines | Gerry Wan, Shinan Liu, Francesco Bronzino, Nick Feamster, Zakir Durumeric | 12m |
CATO automatically searches hundreds of feature + packet-depth configurations, compiles full serving pipelines and measures them end to end, trading off… | ||
| Mutant: Learning Congestion Control from Existing Protocols via Online Reinforcement... | Lorenzo Pappone, Alessio Sacco, Flavio Esposito | 17m |
Mutant is an implementation that runs a kernel protocol manager plus a user-space learner to switch among up to six in-kernel TCP schemes with a 10 ms… | ||
| Resolving Packets from Counters: Enabling Multi-scale Network Traffic Super Resolution... | Xizheng Wang, Libin Liu, Li Chen, Dan Li, Yukai Miao, Yu Bai | 16m |
ZOOMSYNTH reconstructs packet-level traces from second‑level counters by stacking a transformer that models trends with an extreme‑value branch into a tree of… | ||
| BFTBrain: Adaptive BFT Consensus with Reinforcement Learning | Chenyuan Wu, Haoyun Qin, Mohammad Javad Amiri, Boon Thau Loo, Dahlia Malkhi, Ryan Marcus | 17m |
System adapts to live conditions by measuring metered features and using a decentralized deterministic RL agent on each node to agree on protocol changes, then… | ||
| nothing matches | ||