USENIX ATC 2024
2024-07-10..2024-07-12
Santa Clara, United States
as the organisers put it
The 2024 USENIX Annual Technical Conference will take place July 10–12, 2024, at the Hyatt Regency Santa Clara, in Santa Clara, CA, USA. USENIX ATC '24 will bring together leading systems researchers for cutting-edge systems research and the opportunity to gain insight into a wealth of must-know topics.
| title | author | length |
|---|---|---|
| Joint Keynote Address: Scaling AI Sustainably: An Uncharted Territory | Carole-Jean Wu | 01h 02m |
A concrete account of AI’s footprint: operational vs embodied carbon are measured across model life cycles, showing a 2021 translation training cost of 45.2… | ||
| Harmonizing Efficiency and Practicability: Optimizing Resource Utilization in... | Qingyuan Liu, Yanning Yang, Dong Du, Yubin Xia, Ping Zhang, Jia Feng, James R. Larus, Haibo Chen | 21m |
Jiagu targets serverless resource waste with two concrete techniques: pre-decision scheduling builds per-node capacity tables that answer most placement… | ||
| ALPS: An Adaptive Learning, Priority OS Scheduler for Serverless Functions | Yuqi Fu, Ruizhe Shi, Haoliang Wang, Songqing Chen, Yue Cheng | 18m |
ALPS unifies approximate SRPT via user-space SRPT simulation with eBPF hooks to CFS in a decoupled frontend/backend: the user-space front end learns time-slice… | ||
| Starburst: A Cost-aware Scheduler for Hybrid Cloud | Michael Luo, Siyuan Zhuang, Suryaprakash Vengadesan, Romil Bhardwaj, Justin Chang, Eric Friedman, Scott Shenker, Ion Stoica | 19m |
Starburst is a meta-scheduler that assigns per-job waiting budgets, uses compute-weight (or its practical approximation star-weight with preemption) plus… | ||
| StreamBox: A Lightweight GPU SandBox for Serverless Inference Workflow | Hao Wu, Yue Yu, Junxiao Deng, Shadi Ibrahim, Song Wu, Hao Fan, Ziyue Cheng, Hai Jin | 15m |
StreamBox replaces heavy per-function CUDA contexts with concurrent GPU streams plus an auto-scaling memory pool that profiles models offline, does lazy… | ||
| Power-aware Deep Learning Model Serving with μ-Serve | Haoran Qiu, Weichao Mao, Archit Patke, Shengkun Cui, Saurabh Jha, Chen Wang, Hubertus Franke, Zbigniew T. Kalbarczyk, Tamer Başar, Ravishankar K. Iyer | ? |
microserve is a power-aware model serving system that builds a sensitivity profile per model partition, clusters partitions with similar frequency sensitivity… | ||
| Fast Inference for Probabilistic Graphical Models | Jiantong Jiang, Zeyi Wen, Atif Mansoor, Ajmal Mian | 18m |
Fast-PGM is a C++ system for importance-sampling PGM inference that claims 3–20× speedup over prior implementations by fusing CPT and proposal tables, using a… | ||
| Cost-Efficient Large Language Model Serving for Multi-turn Conversations with... | Bin Gao, Zhuomin He, Puru Sharma, Qingxuan Kang, Djordje Jevdjic, Junbo Deng, Xingkun Yang, Zhou Yu, Pengfei Zuo | 22m |
A hierarchical KV caching system called CachedAttention reuses key/value tensors across multi-turn chats by spilling caches to SSD and host DRAM, using… | ||
| PUZZLE: Efficiently Aligning Large Language Models through Light-Weight Context... | Kinman Lei, Yuyang Jin, Mingshu Zhai, Kezhao Huang, Haoxing Ye, Jidong Zhai | 15m |
PUZZLE implements a two-dimensional model-orchestration system that reduces context-switch overhead during RLHF by adding time-sharing within stages and a… | ||
| ScalaAFA: Constructing User-Space All-Flash Array Engine with Holistic Designs | Shushu Yi, Xiurui Pan, Qiao Li, Qiang Li, Chenxi Wang, Bo Mao, Myoungsoo Jung, Jie Zhang | 19m |
ScalaAFA is a user-space all‑flash array engine implemented on SPDK and evaluated with FAMU-emulated PCIe5 SSDs that offload replication‑to‑striping conversion… | ||
| FastCommit: resource-efficient, performant and cost-effective file system... | Harshad Shirwadkar, Saurabh Kadekodi, Theodore Tso | 21m |
Ext4 gains a hybrid journaling mode that replaces many fsync-triggered physical writes with compact logical records, so over 80% of commits fit in a single 4KB… | ||
| ZMS: Zone Abstraction for Mobile Flash Storage | Joo-Young Hwang, Seokhwan Kim, Daejun Park, Yong-Gil Song, Junyoung Han, Seunghyun Choi, Sangyeun Cho, Youjip Won | 20m |
An implemented I/O stack for ZNS on a 128 GB UFS phone uses FTFS, device firmware changes and three techniques IOTailor budget-based in-place update… | ||
| Ethane: An Asymmetric File System for Disaggregated Persistent Memory | Miao Cai, Junru Shen, Baoliu Ye | 18m |
Ethane builds an asymmetric file system for disaggregated persistent memory that splits a control-plane FS (cache FS, dlog/mlog with log Arena to cut… | ||
| PeRF: Preemption-enabled RDMA Framework | Sugi Lee, Mingyu Choi, Ikjun Yeom, Younghoon Kim | 20m |
PeRF implements RNIC preemption in software using managed QPs and zero-weight work requests to break large RDMA transfers into interleaved sub-requests and to… | ||
| CyberStar: Simple, Elastic and Cost-Effective Network Functions Management in... | Tingting Xu, Bengbeng Xue, Yang Song, Xiaomin Wu, Xiaoxin Peng, Yilong Lyu, Xiaoliang Wang, Chen Tian, Baoliu Ye, Camtu Nguyen, Biao Lyu, Rong Wen, Zhigang Zong, Shunmin Zhu | 17m |
Cloud network functions run on ECS (containers/VMs) with a three‑tier SC/PP/fabric architecture that pools resources, shares stateless packet‑processing units… | ||
| OSMOSIS: Enabling Multi-Tenancy in Datacenter SmartNICs | Mikhail Khalilov, Marcin Chrapek, Siyuan Shen, Alessandro Vezzu, Thomas Benz, Salvatore Di Girolamo, Timo Schneider, Daniele De Sensi, Luca Benini, Torsten Hoefler | 17m |
OSMOSIS provides a hardware resource manager that enforces fair multi-tenancy on an open-source RISC‑V 400Gb/s SmartNIC by adding lightweight schedulers for… | ||
| ETC: An Elastic Transmission Control Using End-to-End Available Bandwidth... | Feixue Han, Qing Li, Peng Zhang, Gareth Tyson, Yong Jiang, Mingwei Xu, Yulong Lan, ZhiCheng Li | 20m |
ETC claims a new congestion signal: the pulling rate, measured from the instantaneous receipt rate of N-packet microbursts, and uses a convex elastic step… | ||
| More is Different: Prototyping and Analyzing a New Form of Edge Server with... | Li Zhang, Zhe Fu, Boqing Shi, Xiang Li, Rujin Lai, Chenyang Yang, Ao Zhou, Xiao Ma, Shangguang Wang, Mengwei Xu | 20m |
A commercial SoC Cluster with 60 Snapdragon 865 SoCs in a 2U rack and deployed for mobile cloud gaming is measured against servers using a benchmark suite that… | ||
| HiP4-UPF: Towards High-Performance Comprehensive 5G User Plane Function on P4... | Zhixin Wen, Guanhua Yan | 17m |
HiP4-UPF: Towards High-Performance Comprehensive 5G User Plane Function on P4 Programmable Switches
Zhixin Wen and Guanhua Yan, Binghamton University
Due to… | ||
| KEPC-Push: A Knowledge-Enhanced Proactive Content Push Strategy for... | Ziwen Ye, Qing Li, Chunyu Qiao, Xiaoteng Ma, Yong Jiang, Qian Ma, Shengbin Meng, Zhenhui Yuan, Zili Meng | 21m |
Proposes KEPC-Push, which uses knowledge graph embeddings (TransE, 128‑D vectors) plus collaborative‑filtering/Skip‑gram to infer popularity for fresh videos… | ||
| High-density Mobile Cloud Gaming on Edge SoC Clusters | Li Zhang, Shangguang Wang, Mengwei Xu | 16m |
Claims a system that packs more mobile game sessions onto servers of stacked phone SoCs by doing pre-run game partitioning, pushing rendering work across SoCs… | ||
| Limitations and Opportunities of Modern Hardware Isolation Mechanisms | Xiangdong Chen, Zhaofeng Li, Tirth Jain, Vikram Narayanan, Anton Burtsev | 20m |
Implements an NF virtualization framework (SAR) with a migrating‑thread model, per‑domain private heaps and a shared exchange heap, and evaluates Intel MPK… | ||
| FetchBPF: Customizable Prefetching Policies in Linux with eBPF | Xuechun Cao, Shaurya Patel, Soo Yee Lim, Xueyuan Han, Thomas Pasquier | 11m |
FetchBPF extends eBPF to let developers replace Linux’s sequential prefetcher by loading programs at hook A (collect fault addresses) and hook B (run policy)… | ||
| Fast (Trapless) Kernel Probes Everywhere | Jinghao Jia, Michael V. Le, Salman Ahmed, Dan Williams, Hani Jamjoom, Tianyin Xu | 13m |
Replaces trap-based Kprobes with direct control-flow rewrites and selective padding to avoid double-trap overhead: a compiler LLVM machine IR pass inserts… | ||
| HydraRPC: RPC in the CXL Era | Teng Ma, Zheng Liu, Chengkun Wei, Jialiang Huang, Youwei Zhuo, Haoyu Li, Ning Zhang, Yijin Guan, Dimin Niu, Mingxing Zhang, Tao Ma | 18m |
HydraRPC implements an RPC stack that mounts CXL-attached HDM, exposes 64-bit message-queue entries and preallocated data slots, and removes network round… | ||
| ExtMem: Enabling Application-Aware Virtual Memory Management for Data-Intensive... | Sepehr Jalalian, Shaurya Patel, Milad Rezaei Hajidehi, Margo Seltzer, Alexandra Fedorova | 10m |
ExtMem/XM is a user-space framework that replaces userfaultfd with an upcall-based fault handling path to remove IPC and serialization, lowering page-fault… | ||
| Telescope: Telemetry for Gargantuan Memory Footprint Applications | Alan Nair, Sandeep Kumar, Aravinda Prasad, Ying Huang, Andy Rudoff, Sreenivas Subramoney | 20m |
Telescope is a kernel telemetry technique that profiles higher and lower levels of the radix page table to find hot and cold pages in terabyte-class address… | ||
| An Empirical Study of Rust-for-Linux: The Success, Dissatisfaction, and Compromise | Hongyu Li, Liwei Guo, Yexuan Yang, Shangguang Wang, Mengwei Xu | 20m |
Study analyzes 6 RFL drivers, hundreds of issues/PRs, thousands of commits and 12K Zulip discussions to evaluate Rust-for-Linux after mainline adoption. Core… | ||
| Scalable and Effective Page-table and TLB management on NUMA Systems | Bin Gao, Qingxuan Kang, Hao-Wei Tee, Kyle Timothy Ng Chu, Alireza Sanaee, Djordje Jevdjic | 17m |
Hydra implements on-demand partial replication of page-table entries inside Linux on x86_64, tracking per-node sharers so page-table reads are local while… | ||
| UniMem: Redesigning Disaggregated Memory within A Unified Local-Remote Memory... | Yijie Zhong, Minqiang Zhou, Zhirong Shen, Jiwu Shu | 20m |
UniMem redesigns cache-coherent disaggregated memory by removing the accelerator address-translation hop and treating device-attached and host RAM as a unified… | ||
| WingFuzz: Implementing Continuous Fuzzing for DBMSs | Jie Liang, Zhiyong Wu, Jingzhou Fu, Yiyuan Bai, Qiang Zhang, Yu Jiang | 13m |
WingFuzz implements continuous DBMS fuzzing with a grammar-derived mutator, rapid-commit fuzzing that targets only functions affected by commits, and a… | ||
| Balancing Analysis Time and Bug Detection: Daily Development-friendly Bug... | Keita Suzuki, Kenta Ishiguro, Kenji Kono | 21m |
FiTx framework implements a practical, compiler-integrated checker using only single-compilation-unit analysis, static field offsets, simple aliasing and no… | ||
| Kivi: Verification for Cluster Management | Bingzhe Liu, Gangmuk Lim, Ryan Beckett, P. Brighten Godfrey | 23m |
Kivi verifies Kubernetes controllers and configurations by modeling controllers, objects, and events and exhaustively exploring interleavings with model… | ||
| Monarch: A Fuzzing Framework for Distributed File Systems | Tao Lyu, Liyi Zhang, Zhiyao Feng, Yueyang Pan, Yujie Ren, Meng Xu, Mathias Payer, Sanidhya Kashyap | 22m |
MONARCH builds a push-button fuzzer that tests POSIX DFSes across VMs with deterministic fault injection at syscall granularity, unified cross-node coverage… | ||
| Accelerating the Training of Large Language Models using Efficient Activation... | Tailing Yuan, Yuliang Liu, Xucheng Ye, Shenglong Zhang, Jianchao Tan, Bin Chen, Chengru Song, Di Zhang | 19m |
Accelerating the Training of Large Language Models using Efficient Activation Rematerialization and Optimal Hybrid Parallelism
Tailing Yuan, Yuliang Liu… | ||
| Metis: Fast Automatic Distributed Training on Heterogeneous GPUs | Taegeon Um, Byungsoo Oh, Minyoung Kang, Woo-Yeon Lee, Goeun Kim, Dongseob Kim, Youngtaek Kim, Mohd Muzzammil, Myeongjae Jeon | 20m |
Metis automatically finds parallelism plans for training across mixed GPU types by profiling each GPU, estimating cost, and using a planner that prunes skewed… | ||
| FwdLLM: Efficient Federated Finetuning of Large Language Models with Perturbed... | Mengwei Xu, Dongqi Cai, Yaozong Wu, Xiang Li, Shangguang Wang | 20m |
Introduces a BP-free federated finetuning pipeline that estimates gradients via perturbed inferences (forward gradients) and updates only small adapter… | ||
| A Secure, Fast, and Resource-Efficient Serverless Platform with Function REWIND | Jaehyun Song, Bumsuk Kim, Minwoo Kwak, Byoungyoung Lee, Euiseong Seo, Jinkyu Jeong | 19m |
REWIND claims to eliminate both memory and file persistence for warm containers while keeping latency near baseline by using kernel-level memory snapshots… | ||
| SimEnc: A High-Performance Similarity-Preserving Encryption Approach for... | Tong Sun, Bowen Jiang, Borui Li, Jiamei Lv, Yi Gao, Wei Dong | 18m |
SimEnc builds a prototype in Go and C++ with one GPU and integrates semantic-hash MLE to make similar chunks yield the same encryption key via adaptive… | ||
| mmTLS: Scaling the Performance of Encrypted Network Traffic Inspection | Junghan Yoon, Seunghyun Do, Duckwoo Kim, Taejoong Chung, KyoungSoo Park | 20m |
mmTLS eliminates split-connection relaying by having clients share end-to-end session keys over an out-of-band “autoband” channel so the middlebox decrypts on… | ||
| Pecan: Cost-Efficient ML Data Preprocessing with Automatic Transformation... | Dan Graur, Oto Mraz, Muyu Li, Sepehr Pourghannad, Chandramohan A. Thekkath, Ana Klimovic | 21m |
Pecan is a service that automatically places preprocessing workers on accelerator hosts and applies automatic transformation reordering to cut remote CPU… | ||
| OPER: Optimality-Guided Embedding Table Parallelization for Large-scale... | Zheng Wang, Yuke Wang, Boyuan Feng, Guyue Huang, Dheevatsa Mudigere, Bharath Muthiah, Ang Li, Yufei Ding | 20m |
Claims a near-optimal planner and system that cut end-to-end DLRM embedding lookup cost using a heuristic search plus a distributed shared memory runtime and a… | ||
| MAGPY: Compiling Eager Mode DNN Programs by Monitoring Execution States | Chen Zhang, Rongchao Dong, Haojie Wang, Runxin Zhong, Jike Chen, Jidong Zhai | 19m |
MAGPY monitors the Python interpreter’s execution states to record a reference graph of object references and versions, generates guards and small mo (mock)… | ||
| Quant-LLM: Accelerating the Serving of Large Language Models via FP6-Centric... | Haojun Xia, Zhen Zheng, Xiaoxia Wu, Shiyang Chen, Zhewei Yao, Stephen Youn, Arash Bakhtiari, Michael Wyatt, Donglin Zhuang, Zhongzhu Zhou, Olatunji Ruwase, Yuxiong He, Shuaiwen Leon Song | 19m |
Quant-LLM demonstrates that FP6 weight quantization preserves LLM quality while enabling much cheaper serving by building a CUDA kernel suite called TC-FPx… | ||
| QDSR: Accelerating Layer-7 Load Balancing by Direct Server Return with QUIC | Ziqi Wei, Zhiqiang Wang, Qing Li, Yuan Yang, Cheng Luo, Fuyu Wang, Yong Jiang, Sijie Yang, Zhenhui Yuan | 17m |
QDSR uses QUIC plus Direct Server Return to split a QUIC connection into independent streams and perform a two-phase stream handoff so multiple real servers… | ||
| Evaluating Chiplet-based Large-Scale Interconnection Networks via Cycle-Accurate... | Yinxiao Feng, Yuchen Wei, Dong Xiang, Kaisheng Ma | 21m |
A cycle-accurate packet-parallel simulator called CNSim implements an atomic-based hyper-threading model and packet-centric architecture to evaluate… | ||
| Config-Snob: Tuning for the Best Configurations of Networking Protocol Stack | Manaf Bin-Yahya, Yifei Zhao, Hossein Shafieirad, Anthony Ho, Shijun Yin, Fanzhao Wang, Geng Li | 16m |
Config-Snob selects configurations from an eight-knob protocol space using historical data, two-level client grouping, and a multi-arm bandit that bootstraps a… | ||
| Conspirator: SmartNIC-Aided Control Plane for Distributed ML Workloads | Yunming Xiao, Diman Zad Tootaghaj, Aditya Dhakal, Lianjie Cao, Puneet Sharma, Aleksandar Kuzmanovic | 18m |
Conspirator is a SmartNIC-based control plane that offloads data movement and scheduling to an off-path BlueField SmartNIC, using DMA to push client tensors… | ||
| FBMM: Making Memory Management Extensible With Filesystems | Bijan Tabatabai, James Sorenson, Michael M. Swift | 18m |
Proposes using Linux VFS callbacks to implement memory policies as FBMM file systems so memory is allocated by creating and mapping files in a mount directory… | ||
| Mangosteen: Fast Transparent Durability for Linearizable Applications using NVM | Sergey Egorov, Gregory Chockler, Brijesh Dongol, Dan O'Keeffe, Sadegh Keshavarzi | 22m |
Mangosteen converts a linearizable in-memory app to durably linearizable NVM-backed code using six callback hooks, runtime binary instrumentation for… | ||
| FlexMem: Adaptive Page Profiling and Migration for Tiered Memory | Dong Xu, Junhee Ryu, Jinho Baek, Kwangsik Shin, Pengfei Su, Dong Li | 17m |
FlexMem is a page-profiling and migration system that combines performance-counter and fault-hint profiling to catch both long-lived and emerging hot pages… | ||
| SuperBench: Improving Cloud AI Infrastructure Reliability with Proactive Validation | YIFAN XIONG, Yuting Jiang, Ziyue Yang, Lei Qu, Guoshuai Zhao, Shuguang Liu, Dong Zhong, Boris Pinzur, Jie Zhang, Yang Wang, Jithin Jose, Hossein Pourreza, Jeff Baxter, Kushal Datta, Prabhat Ram, Luke Melton, Joe Chau, Peng Cheng, Yongqiang Xiong, Lidong Zhou | 18m |
SuperBench is a proactive validation system that runs a comprehensive benchmark suite (end‑to‑end and microbenchmarks), uses a Validator that learns benchmark… | ||
| Removing Obstacles before Breaking Through the Memory Wall: A Close Look at HBM... | Ronglong Wu, Shuyue Zhou, Jiahao Lu, Zhirong Shen, Zikang Xu, Jiwu Shu, Kunlin Yang, Feilong Lin, Yiming Zhang | 21m |
Study analyzes >460 million HBM error events from ~60,000 chips across 19 datacenters over two years and finds HBM-specific patterns: concentrate in 7th, 15th… | ||
| MSFRD: Mutation Similarity based SSD Failure Rating and Diagnosis for Complex... | Yuqi Zhang, Tianyi Zhang, Wenwen Hao, Shuyang Wang, Na Liu, Xing He, Yang Zhang, Weixin Wang, Yongguang Cheng, Huan Wang, Jie Xu, Feng Wang, Bo Jiang, Yongwong Gwon, JongSung Na, Zoe Kim, Geunrok Oh | 20m |
Proposes MSFRD, which uses dynamic mutation extraction from telemetry (51M+ T logs, 85 custom attributes) and a weighted mutation-similarity metric to score… | ||
| Diagnosing Application-network Anomalies for Millions of IPs in Production Clouds | Zhe Wang, Huanwu Hu, Linghe Kong, Xinlei Kang, Teng Ma, Qiao Xiang, Jingxuan Li, Yang Lu, Zhuo Song, Peihao Yang, Jiejian Wu, Yong Yang, Tao Ma, Zheng Liu, Xianlong Zeng, Dennis Cai, Guihai Chen | 21m |
AND is a deployed Alibaba Cloud system that diagnoses application-network anomalies for millions of IPs by using TCP retransmissions as the single signal… | ||
| Data Caching for Enterprise-Grade Petabyte-Scale OLAP | Chunxu Tang, Bin Fan, Jing Zhao, Chen Liang, Yi Wang, Beinan Wang, Ziyue Qiu, Lu Qiu, Bowen Ding, Shouzhuo Sun, Saiguang Che, Jiaming Mai, Shouwei Chen, Yu Zhu, Jianjian Xie, Yutian (James) Sun, Yao Li, Yangjun Zhang, Ke Wang, Mingmin Chen | 20m |
Alluxio local cache embeds on-node SSDs and a page store to serve reads for Presto and HDFS, cutting network I/O and API calls. Deployed at Uber and Meta for… | ||
| Full Lifecycle Data Analysis on a Large-scale and Leadership Supercomputer... | Bin Yang, Hao Wei, Wenhao Zhu, Yuhao Zhang, Weiguo Liu, Wei Xue | 20m |
Six years of 40 TB over six years from Sunway TaihuLight (41,508 nodes) and bacon monitoring (deployed April 2017) analyze 6 million jobs totaling 4 billion… | ||
| Panorama: Optimizing Internet-scale Users’ Routes from End to End | Geng Li, Shuihai Hu, Kun Tan | 21m |
Panorama is a globally deployed hybrid overlay that runs a centralized controller and distributed route servers to compute end-to-end routes using… | ||
| Enhancing Resource Management of the World's Largest PCDN System for On-Demand... | Rui-Xiao Zhang, Haiping Wang, Shu Shi, Xiaofei Pang, Yajie Peng, Zhichen Xue, Jiangchuan Liu | 19m |
ByteDance documents a deployed system called PCDN with 1M nodes in 200+ cities where 80% of nodes deliver <5 Mbps, 23% sit behind nonpublic IPs and failures… | ||
| TileClipper: Lightweight Selection of Regions of Interest from Videos... | Shubham Chaudhary, Aryan Taneja, Anjali Singh, Purbasha Roy, Sohum Sikdar, Mukulika Maity, Arani Bhattacharya | 15m |
TileClipper selects rectangular tiles from H.265-encoded traffic video by using tile bitrate as a noisy proxy for object count, calibrating thresholds by… | ||
| Expeditious High-Concurrency MicroVM SnapStart in Persistent Memory with an... | Xingguo Pang, Yanze Zhang, Liu Liu, Dazhao Cheng, Chengzhong Xu, Xiaobo Zhou | 20m |
PASS is a user-space augmented hypervisor that maps entire MicroVM snapshots into a single-tier Intel PMEM address space and implements zero-copy on-demand… | ||
| Taming Hot Bloat Under Virtualization with HUGESCOPE | Chuandong Li, Sai Sha, Yangqing Zeng, Xiran Yang, Yingwei Luo, Xiaolin Wang, Zhenlin Wang, Diyu Zhou | 17m |
HugeScope is a hypervisor-side system that detects and fixes “hot bloat” by two-phase page tracking (coarse scan then fine-grained base-page tracking) while… | ||
| CrossMapping: Harmonizing Memory Consistency in Cross-ISA Binary Translation | Chen Gao, Xiangwei Meng, Wei Li, Jinhui Lai, Yiran Zhang, Fengyuan Ren | 17m |
CrossMapping is a framework that generates concrete fence- and RMW-mapping schemes for cross-ISA DBT, implemented on QEMU and evaluated with PARSEC. It uses… | ||
| Efficient Decentralized Federated Singular Vector Decomposition | Di Chai, Junxue Zhang, Liu Yang, Yilun Jin, Leye Wang, Kai Chen, Qiang Yang | 19m |
Excalibur is a decentralized federated SVD system that replaces external servers with a multiplicative matrix sharing scheme and a communication-optimized… | ||
| Models on the Move: Towards Feasible Embedded AI for Intrusion Detection on... | He Xu, Di Wu, Yufeng Lu, Jiwu Lu, Haibo Zeng | 18m |
Proposes MULSAM, a multi-dimensional LSTM with self-attention tuned for vehicular CAN bus intrusion detection, trained on five attack types synthesized from… | ||
| CPC: Flexible, Secure, and Efficient CVM Maintenance with Confidential Procedure... | Jiahao Chen, Zeyu Mi, Yubin Xia, Haibing Guan, Haibo Chen | 20m |
CPC (Confidential Procedure Calls) runs vetted maintenance code inside CVMs using extra host‑wakeup vCPU threads (hvvCPUs) and a state‑machine API, avoiding… | ||
| RL-Watchdog: A Fast and Predictable SSD Liveness Watchdog on Storage Systems | Jin Yong Ha, Sangjin Lee, Heon Y. Yeom, Yongseok Son | 18m |
A kernel-level watchdog implements a LWW liveness probe, a reinforcement-learning timeout predictor (RLTP) and a fast-failure-notify path (FFN) inside Linux… | ||
| Exploit both SMART Attributes and NAND Flash Wear Characteristics to Effectively... | Yunfei Gu, Chentao Wu, Xubin He | 21m |
Claims device-level NAND wear histograms plus SMART logs improve SSD failure forecasts, using manufacturer-extracted wear tests and 20,000+ self-test logs to… | ||
| StreamCache: Revisiting Page Cache for File Scanning on Fast Storage Devices | Zhiyue Li, Guangyan Zhang | 21m |
Revisits kernel page cache for scan workloads and builds StreamCache in XFS with a stream tracking tree, stream-granular reclaim, and a two-layer memory… | ||
| Scalable Billion-point Approximate Nearest Neighbor Search Using SmartSSDs | Bing Tian, Haikun Liu, Zhuohui Duan, Xiaofei Liao, Hai Jin, Yu Zhang | 18m |
SmartANNS implements hierarchical indices on host plus Samsung SmartSSDs with FPGA acceleration, dynamic task scheduling and a gradient-boosted shard pruning… | ||
| gVulkan: Scalable GPU Pooling for Pixel-Grained Rendering in Ray Tracing | Yicheng Gu, Yun Wang, Yunfan Sun, Yuxin Xiang, Yufan Jiang, Xuyan Hu, Zhengwei Qi, Haibing Guan | 18m |
gVulkan implements transparent API-forwarding for Vulkan to offload ray tracing to a cloud GPU pool, uses resource-classified API-forwarding to avoid round… | ||
| vFPIO: A Virtual I/O Abstraction for FPGA-accelerated I/O Devices | Jiyang Chen, Harshavardhan Unnibhavi, Atsushi Koshiba, Pramod Bhatotia | 20m |
vFPIO implements a virtual I/O module compiled into user logic, a vFPIO shell with a preemptive I/O transaction scheduler, and a host library with POSIX-like… | ||
| ScalaCache: Scalable User-Space Page Cache Management with Software-Hardware... | Li Peng, Yuda An, You Zhou, Chenxi Wang, Qiao Li, Chuanning Cheng, Jie Zhang | 17m |
ScalaCache moves page‑cache management into computational storage drives (CSDs), implements a fusion FTL that merges indirection layers, and runs lock‑free… | ||
| Centimani: Enabling Fast AI Accelerator Selection for DNN Training with a Novel... | Zhen Xie, Murali Emani, Xiaodong Yu, Dingwen Tao, Xin He, Pengfei Su, Keren Zhou, Venkatram Vishwanath | 22m |
Centimani builds a memory estimation model plus decoupled performance models and uses microbenchmarks and roofline-style measurements to pick batch size and… | ||
| A Difference World: High-performance, NVM-invariant, Software-only Intermittent... | Harrison Williams, Saim Ahmad, Matthew Hicks | 17m |
Camel is a compiler+runtime that makes software-only intermittent computation practical on Flash devices by splitting scarce SRAM into two mixed-volatility… | ||
| Efficient Large Graph Processing with Chunk-Based Graph Representation Model | Rui Wang, Weixu Zong, Shuibing He, Xinyu Chen, Zhenxin Li, Zheng Dang | 09m |
ChunkGraph implements a chunk-based graph representation on NVMe SSDs, classifying vertices into tiny, medium and super categories and using a four-layer… | ||
| SlimArchive: A Lightweight Architecture for Ethereum Archive Nodes | Hang Feng, Yufeng Hu, Yinghan Kou, Runhuai Li, Jianfeng Zhu, Lei Wu, Yajin (Andy) Zhou | 20m |
Prototype SLIMARCHIVE implements a recorder, encoder and state generator to store flattened, transaction‑level deltas in a state temporal archive instead of… | ||
| Every Mapping Counts in Large Amounts: Folio Accounting | David Hildenbrand, Martin Schulz, Nadav Amit | 12m |
Implements a Linux folio accounting scheme that uses the Pigeonhole principle plus a compressed per-folio address-space aggregator derived from small… | ||
| nothing matches | ||