USENIX ATC '20
2020-07-15..2020-07-17
-- online --
as the organisers put it
2020 USENIX Annual Technical Conference will take place July 15–17, 2020. USENIX ATC '20 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 |
|---|---|---|
| Libnvmmio:Reconstructing Software IO Path w/ Failure-Atomic Memory-Mapped Interface | Jungsik Choi, Jaewan Hong, Youngjin Kwon, Hwansoo Han | 22m |
Libnvmmio: Reconstructing Software IO Path with Failure-Atomic Memory-Mapped Interface
Jungsik Choi, Sungkyunkwan University; Jaewan Hong and Youngjin Kwon… | ||
| MatrixKV: Reducing Write Stalls and Write Amplification in LSM-tree Based KV Stores | Ting Yao, Yiwen Zhang, Jiguang Wan, Qiu Cui, Liu Tang, Hong Jiang, ChangSheng Xie, Xubin He | 21m |
MatrixKV implements a matrix container in NVM (Optane DC PMM) inside RocksDB, adds column compaction and cross-row hint search, and flattens levels to cut… | ||
| Disaggregating Persistent Memory and Controlling Them Remotely | Shin-Yeh Tsai, Yizhou Shan, Yiying Zhang | 25m |
Explores a passive disaggregated persistent memory model (pDPM), implements three remote key-value systems—pDPM-direct, pDPM-central, and Clover—and measures… | ||
| SplinterDB: Closing the Bandwidth Gap for NVMe Key-Value Stores | Alex Conway, Abhishek Gupta, Vijay Chidambaram, Martin Farach-Colton, Richard Spillane, Amy Tai, Rob Johnson | 24m |
SplinterDB implements a new data structure, the STBε-tree, plus a flush-and-compact execution model and a small-cache design, and it runs on an Intel Optane… | ||
| Twizzler: a Data-Centric OS for Non-Volatile Memory | Daniel Bittman, Peter Alvaro, Pankaj Mehra, Darrell D. E. Long, Ethan L. Miller | 24m |
Twizzler redesigns the OS for byte-addressable NVM by giving programs memory-style access to persistent data using 64-bit object-relative pointers backed by… | ||
| BASTION: A Security Enforcement Network Stack for Container Networks | Jaehyun Nam, Seungsoo Lee, Hyunmin Seo, Phil Porras, Vinod Yegneswaran, Seungwon Shin | 22m |
BASTION is a container-side network stack that enforces per-container topology and point-to-point traffic isolation, implementing a network visibility service… | ||
| Spool: Reliable Virtualized NVMe Storage Pool in Public Cloud Infrastructure | Shuai Xue, Shang Zhao, Quan Chen, Gang Deng, Zheng Liu, Jie Zhang, Zhuo Song, Tao Ma, Yong Yang, Yanbo Zhou, Keqiang Niu, Sijie Sun, Minyi Guo | 21m |
Spool is a NVMe virtualization system built on SPDK and a user-space NVMe driver that diagnoses device-failure type and resets controllers for link errors… | ||
| HDDse: Enabling High-Dimensional Disk State Embedding for Generic Failure Detection | Ji Zhang, Ping Huang, Ke Zhou, Ming Xie, Sebastian Schelter | 20m |
HDDse uses an LSTM-based siamese network to map daily SMART snapshots into unified high-dimensional disk state embeddings and classifies similarity pairs with… | ||
| Adaptive Placement for In-memory Storage Functions | Ankit Bhardwaj, Chinmay Kulkarni, Ryan Stutsman | 26m |
A system implements a server-internal cost model that observes user-defined functions and dynamically chooses client or storage execution for each invocation… | ||
| NetKernel: Making Network Stack Part of the Virtualized Infrastructure | Zhixiong Niu, Hong Xu, Peng Cheng, Qiang Su, Yongqiang Xiong, Tao Wang, Dongsu Han, Keith Winstein | 20m |
NetKernel decouples the TCP/IP stack into an operator-run network stack module implemented as a VM, redirecting BSD socket calls via a guest shim into… | ||
| Platinum: A CPU-Efficient Concurrent Garbage Collector | Mingyu Wu, Ziming Zhao, Yanfei Yang, Haoyu Li, Haibo Chen, Binyu Zang, Haibing Guan, Sanhong Li, Chuansheng Lu, Tongbao Zhang | 19m |
Platinum implements a mostly-concurrent JVM GC that isolates mutators and GC threads by binding GC threads to cores and moving mutators to idle-core pinning… | ||
| PinK: High-speed In-storage Key-value Store with Bounded Tails | Junsu Im, Jinwook Bae, Chanwoo Chung, Sungjin Lee | 22m |
PinK implements an LSM-tree KV-SSD that pins the top levels of the tree in DRAM to bound reads to a single flash access, removes Bloom filters to cut CPU work… | ||
| OPTIMUSCLOUD: Heterogeneous Configuration Optimization for Distributed Databases | Ashraf Mahgoub, Alexander Michaelson Medoff, Rakesh Kumar, Subrata Mitra, Ana Klimovic, Somali Chaterji, Saurabh Bagchi | 22m |
System builds an online reconfiguration engine that jointly tunes NoSQL parameters and EC2 VM types, partitions clusters into complete sets to limit… | ||
| Serverless in the Wild: Characterizing and Optimizing the Serverless Workload | Mohammad Shahrad, Rodrigo Fonseca, Íñigo Goiri, Gohar Irfan Chaudhry, Paul Batum, Jason Cooke, Eduardo Laureano, Colby Tresness, Mark Russinovich, Ricardo Bianchini | 23m |
Two weeks of July 2019 Azure Functions traces show a highly skewed workload: 82% apps <1/min account for only 0.4% of invocations but, if kept always warm… | ||
| Lessons Learned from the Chameleon Testbed | Kate Keahey, Jason Anderson, Zhuo Zhen, Pierre Riteau, Paul Ruth, Dan Stanzione, Mert Cevik, Jacob Colleran, Haryadi S. Gunawi, Cody Hammock, Joe Mambretti, Alexander Barnes, François Halbah, Alex Rocha, Joe Stubbs | 25m |
Chameleon adapts mainstream cloud tech into a research-focused platform: it is an openstack-based cloud with deep reconfigurability, advanced reservations via… | ||
| SPINFER: Inferring Semantic Patches for the Linux Kernel | Lucas Serrano, Van-Anh Nguyen, Ferdian Thung, Lingxiao Jiang, David Lo, Julia Lawall, Gilles Muller | 21m |
Spinfer is a tool that infers Coccinelle semantic patches from example edits and outputs human-checkable rules that Coccinelle applies as concrete diffs. It… | ||
| FuZZan: Efficient Sanitizer Metadata Design for Fuzzing | Yuseok Jeon, WookHyun Han, Nathan Burow, Mathias Payer | 21m |
FuZZan replaces ASan’s heavy shadow memory with two compact metadata layouts (mean-shadow and arbitrary metadata), adds a runtime dynamic metadata switching… | ||
| PracExtractor: Extracting Configuration Good Practices from Manuals to Detect... | Chengcheng Xiang, Haochen Huang, Andrew Yoo, Yuanyuan Zhou, Shankar Pasupathy | 20m |
PracExtractor uses NLP (keyword + syntactic pattern filtering, dependency parsing, entity identification and semantic pattern matching) to turn manual… | ||
| Reverse Debugging of Kernel Failures in Deployed Systems | Xinyang Ge, Ben Niu, Weidong Cui | 16m |
Reverse Debugging of Kernel Failures in Deployed Systems
Xinyang Ge, Microsoft Research; Ben Niu, Microsoft; Weidong Cui, Microsoft Research
Post-mortem… | ||
| Offload Annotations: Bringing Heterogeneous Computing to Existing Libraries... | Gina Yuan, Shoumik Palkar, Deepak Narayanan, Matei Zaharia | 23m |
Offload Annotations maps existing CPU libraries like NumPy, Pandas and scikit‑learn to GPU kernels (PyTorch, cuML) via small annotation files and a runtime… | ||
| The Future of the Past: Challenges in Archival Storage | Ethan L. Miller | 01h 19m |
Argues that preserving bits and their meaning for centuries requires choosing media by concrete tradeoffs: tape (5–100 s load, ~200+ MB/s, ~$5/TB) and optical… | ||
| HetPipe: Enabling Large DNN Training on (Whimpy) Heterogeneous GPU Clusters through | Jay H. Park, Gyeongchan Yun, Chang M. Yi, Nguyen T. Nguyen, Seungmin Lee, Jaesik Choi, Sam H. Noh, Young-ri Choi | 20m |
HetPipe is a system that combines pipelined model parallelism with data parallelism by grouping heterogeneous GPUs into a virtual worker that runs concurrent… | ||
| AutoSys: The Design and Operation of Learning-Augmented Systems | Chieh-Jan Mike Liang, Hui Xue, Mao Yang, Lidong Zhou, Lifei Zhu, Zhao Lucis Li, Zibo Wang, Qi Chen, Quanlu Zhang, Chuanjie Liu, Wenjun Dai | 23m |
AutoSys is a unified framework that runs Hyperband, TPE, SMAC tuners and trial benchmarks to auto-tune five Web Search engines (selection, ranking, KV stores)… | ||
| Daydream: Accurately Estimating the Efficacy of Optimizations for DNN Training | Hongyu Zhu, Amar Phanishayee, Gennady Pekhimenko | 25m |
Daydream builds a fine-grained dependency graph from CUPTI traces plus per-layer timestamps, applies simple graph-transformation primitives (select, shrink… | ||
| ALERT: Accurate Learning for Energy and Timeliness | Chengcheng Wan, Muhammad Santriaji, Eri Rogers, Henry Hoffmann, Michael Maire, Shan Lu | 20m |
ALERT is a runtime scheduler that simultaneously picks a DNN and a power cap per input, using a probabilistic global slowdown factor and a Kalman filter to… | ||
| NeuOS: A Latency-Predictable Multi-Dimensional Optimization Framework for DNN-drive | Soroush Bateni, Cong Liu | 26m |
NeuOS implements coordinated layer-level DVFS and low-rank approximation with precomputed hashed speedup-power tables to schedule multiple DNNs on Jetson TX2… | ||
| PERCIVAL: Making In-Browser Perceptual Ad Blocking Practical with Deep Learning | Zainul Abi Din, Panagiotis Tigas, Samuel T. King, Benjamin Livshits | 21m |
PERCIVAL is an in-browser deep learning ad blocker that hooks into Chromium/Brave at the image decode/raster chokepoint to classify every decoded image and… | ||
| Harmonizing Performance and Isolation in Microkernels with Efficient Intra-kernel | Jinyu Gu, Xinyue Wu, Wentai Li, Nian Liu, Zeyu Mi, Yubin Xia, Haibo Chen | 21m |
UnderBridge redesigns microkernels by moving system servers into kernel space while enforcing isolation with Intel PKU/MPK, implementing a fast IPC gate that… | ||
| Faasm: Lightweight Isolation for Efficient Stateful Serverless Computing | Simon Shillaker, Peter Pietzuch | 23m |
Faasm implements thread-based Faaslets on WebAssembly SFI to let co‑located functions share memory directly, use a minimal POSIX host interface, and restore… | ||
| Fewer Cores, More Hertz: Leveraging High-Frequency Cores in the OS Scheduler for | Redha Gouicem, Damien Carver, Jean-Pierre Lozi, Julien Sopena, Baptiste Lepers, Willy Zwaenepoel, Nicolas Palix, Julia Lawall, Gilles Muller | 22m |
Identifies a concrete problem—frequency inversion—caused by per‑core DVFS and long frequency transition latencies (measured 29 ms rise and 98 ms fall)… | ||
| vSMT-IO: Improving I/O Performance and Efficiency on SMT Processors in Virtualized | Weiwei Jia, Jianchen Shan, Tsz On Li, Xiaowei Shang, Heming Cui, Xiaoning Ding | 20m |
Proposes context retention on SMT cores: a hardware thread holds a vCPU context while it waits for I/O so the system avoids busy polling and vCPU switches… | ||
| Lightweight Preemptible Functions | Sol Boucher, Anuj Kalia, David G. Andersen, Michael Kaminsky | 26m |
Introduces a synchronous function-call timeout that runs preemptible code on the same kernel thread, enabling pause/resume/cancel without processes or kernel… | ||
| coIOMMU: A Virtual IOMMU with Cooperative DMA Buffer Tracking for Efficient Memory | Kun Tian, Yu Zhang, Luwei Kang, Yan Zhao, Yaozu Dong | 25m |
coIOMMU implements a virtual IOMMU in QEMU/KVM with a shared DMA Tracking Table (DTT), a pmanager, and a remap engine, and shows that cooperative tracking cuts… | ||
| BatchCrypt: Efficient Homomorphic Encryption for Cross-Silo Federated Learning | Chengliang Zhang, Suyi Li, Junzhe Xia, Wei Wang, Feng Yan, Yang Liu | 23m |
BatchCrypt encodes batches of quantized gradients into single long integers, uses two's-complement packing plus zeros padding and an analytical clipping method… | ||
| A Deep Dive into DNS Query Failures | Donghui Yang, Zhenyu Li, Gareth Tyson | 20m |
Passive DNS logs cover 3.1 billion queries and reveal a 13.5% failure rate overall. A vs AAAA contrast is stark: A queries mostly succeed while AAAA succeed… | ||
| A Decentralized Blockchain with High Throughput and Fast Confirmation | Chenxing Li, Peilun Li, Dong Zhou, Zhe Yang, Ming Wu, Guang Yang, Wei Xu, Fan Long, Andrew Chi-Chih Yao | 22m |
Conflux implements a production Rust blockchain and EVM on Amazon EC2 and demonstrates on up to 12,000 full nodes that the novel Tree-Graph ledger plus the… | ||
| Reconstructing proprietary video streaming algorithms | Maximilian Grüner, Melissa Licciardello, Ankit Singla | 20m |
Reconstructs proprietary ABR logic from player and network logs across roughly 9 hours of traces, learning interpretable concise decision trees over about… | ||
| Midgress-aware traffic provisioning for content delivery | Aditya Sundarrajan, Mangesh Kasbekar, Ramesh K. Sitaraman, Samta Shukla | 18m |
Proposes midgress-aware traffic provisioning that assigns hundreds of traffic classes to CDN servers to cut cache-miss (midgress) traffic; uses footprint… | ||
| GraphWalker: An I/O-Efficient and Resource-Friendly Graph Analytic System... | Rui Wang, Yongkun Li, Hong Xie, Yinlong Xu, John C. S. Lui | 20m |
System design claims an state-aware I/O model plus asynchronous walk updating and block-centric indexing that let a single 64 GB machine with RAID‑0 SSDs… | ||
| Scaph: Scalable GPU-Accelerated Graph Processing with Value-Driven Differential... | Long Zheng, Xianliang Li, Yaohui Zheng, Yu Huang, Xiaofei Liao, Hai Jin, Jingling Xue, Zhiyuan Shao, Qiang-Sheng Hua | 23m |
Scaph is a GPU-accelerated system that partitions graphs into host-resident subgraphs and adaptively classifies each iteration whether a subgraph is high-value… | ||
| Peregreen – modular database for efficient storage of historical time series... | Alexander Visheratin, Alexey Struckov, Semen Yufa, Alexey Muratov, Denis Nasonov, Nikolay Butakov, Yury Kuznetsov, Michael May | 18m |
Peregreen is a modular distributed time-series database that stores all data on Amazon S3 and minimizes backend I/O with a three-tier index, per-sensor… | ||
| AC-Key: Adaptive Caching for LSM-based Key-Value Stores | Fenggang Wu, Ming-Hong Yang, Baoquan Zhang, David H.C. Du | 25m |
AC-Key implements an adaptive cache inside RocksDB that manages key-value cache, key-pointer cache, and block cache, resizing them via hierarchical ghost… | ||
| POSH: A Data-Aware Shell | Deepti Raghavan, Sadjad Fouladi, Philip Levis, Matei Zaharia | 22m |
POSH is a drop‑in shell that offloads I/O‑heavy pipeline stages to proxy servers colocated with NFS storage, using an annotation language to declare which… | ||
| The Fine Line between Bold and Fringe Lunatic | Margo Seltzer | 01h 16m |
Argues that whole-system provenance (PASS) actually captures useful, practical evidence beyond database semantics, shows a hacked Linux kernel (PASS v1) that… | ||
| FineStream: Fine-Grained Window-Based Stream Processing on CPU-GPU Integrated... | Feng Zhang, Lin Yang, Shuhao Zhang, Bingsheng He, Wei Lu, Xiaoyong Du | 25m |
FineStream builds a windowed SQL stream engine for CPU-GPU integrated architectures that uses online profiling and a performance model to do fine-grained… | ||
| OpenExpress: Fully Hardware Automated Open Research Framework for Future Fast... | Myoungsoo Jung | 25m |
OpenExpress is an open RTL framework that implements NVMe queue, PRP parsing and MSI handling entirely in logic and runs on an UltraScale FPGA with 4×64 GB… | ||
| Fast Software Cache Design for Network Appliances | Dong Zhou, Huacheng Yu, Michael Kaminsky, David G. Andersen | 23m |
Replaces Open vSwitch’s microflow table with a new table layout and eviction policy: Bounded Linear Probing lets keys spill into the next bucket to boost… | ||
| Reexamining Direct Cache Access to Optimize I/O Intensive Applications for Multi | Alireza Farshin, Amir Roozbeh, Gerald Q. Maguire Jr., Dejan Kostić | 21m |
Microbenchmarks on two servers with 100 Gbps NICs show DDIO (Intel DDIO) causes about a 30% increase in 99th‑percentile latency when aggregate receive rate… | ||
| sRDMA -- Efficient NIC-based Authentication and Encryption for Remote Direct Memory | Konstantin Taranov, Benjamin Rothenberger, Adrian Perrig, Torsten Hoefler | 27m |
Proposes sRDMA: symmetric‑crypto RDMA with cryptographic work moved onto the NIC, implemented on a Broadcom Stingray ePS225 smartNIC (8‑core ARM, 8 GB DRAM)… | ||
| UREQA: Leveraging Operation-Aware Error Rates for Effective Quantum Circuit Mapping | Tirthak Patel, Baolin Li, Rohan Basu Roy, Devesh Tiwari | 26m |
UREQA builds operation-aware k-NN models (single-qubit, two-qubit, readout) trained from thousands of microbenchmarks plus calibration features (T1, T2… | ||
| Austere Flash Caching with Deduplication and Compression | Qiuping Wang, Jinhong Li, Wen Xia, Erik Kruus, Biplob Debnath, Patrick P. C. Lee | 20m |
AustereCache claims to cut in-memory index overhead by 69.9–97.0% versus prior dedup+compress flash caches while keeping similar read hit and write-reduction… | ||
| DADI: Block-Level Image Service for Agile and Elastic Application Deployment | Huiba Li, Yifan Yuan, Rui Du, Kai Ma, Lanzheng Liu, Windsor Hsu | 21m |
A production block-based image service that replaces download+unpack with on‑demand reads, using a block-level image format plus an overlay block device and a… | ||
| Efficient Miss Ratio Curve Computation for Heterogeneous Content Popularity | Damiano Carra, Giovanni Neglia | 23m |
The paper shows that sampling-based MRC reconstruction breaks when a few very popular items dominate, producing large head errors, and it builds a hybrid… | ||
| Can Applications Recover from fsync Failures? | Anthony Rebello, Yuvraj Patel, Ramnatthan Alagappan, Andrea C. Arpaci-Dusseau, Remzi H. Arpaci-Dusseau | 23m |
The study injects deterministic block failures on an unmodified kernel using dm-loki fault injector and a user-space tester cuttlefs user-space page cache and… | ||
| DupHunter: Flexible High-Performance Deduplication for Docker Registries | Nannan Zhao, Hadeel Albahar, Subil Abraham, Keren Chen, Vasily Tarasov, Dimitrios Skourtis, Lukas Rupprecht, Ali Anwar, Ali R. Butt | 22m |
DupHunter is a Docker registry architecture that natively deduplicates unpacked layer files, stores unique file chunks across servers as layer slices and slice… | ||
| OSCA: An Online-Model Based Cache Allocation Scheme in Cloud Block Storage Systems | Yu Zhang, Ping Huang, Ke Zhou, Hua Wang, Jianying Hu, Yongguang Ji, Bin Cheng | 22m |
OSCA presents an online cache allocator that builds miss-ratio curves from a low-overhead re-access ratio O(1) model (no trace collection), converts those MRCs… | ||
| Lock-free Concurrent Level Hashing for Persistent Memory | Zhangyu Chen, Yu Hua, Bo Ding, Pengfei Zuo | 23m |
Lock-free Concurrent Level Hashing for Persistent Memory
Zhangyu Chen, Yu Hua, Bo Ding, and Pengfei Zuo, Huazhong University of Science and Technology
With… | ||
| Optimizing Memory-mapped I/O for Fast Storage Devices | Anastasios Papagiannis, Giorgos Xanthakis, Giorgos Saloustros, Manolis Marazakis, Angelos Bilas | 20m |
Optimizing Memory-mapped I/O for Fast Storage Devices
Anastasios Papagiannis, Giorgos Xanthakis, Giorgos Saloustros, Manolis Marazakis, and Angelos Bilas… | ||
| A Comprehensive Analysis of Superpage Management Mechanisms and Policies | Weixi Zhu, Alan L. Cox, Scott Rixner | 20m |
Quicksilver implements reservation-based speculative allocation on FreeBSD, combining hybrid incremental-then-full preparation (promote on a utilization… | ||
| Effectively Prefetching Remote Memory with Leap | Hasan Al Maruf, Mosharaf Chowdhury | 22m |
Leap is a Linux kernel patch (tested on 4.4) that implements an online, majority-based prefetching algorithm plus a stripped data path that cuts out generic… | ||
| go-pmem: Native Support for Programming Persistent Memory in Go | Jerrin Shaji George, Mohit Verma, Rajesh Venkatasubramanian, Pratap Subrahmanyam | 24m |
go-pmem is an open-source extension to the Go 1.11 compiler and runtime that adds a garbage-collected persistent heap, native pointers with pointer swizzling… | ||
| End the Senseless Killing: Improving Memory Management for Mobile Operating Systems | Niel Lebeck, Arvind Krishnamurthy, Henry M. Levy, Irene Zhang | 18m |
Marvin implements a runtime-and-OS co-designed memory manager for Android that reintroduces swapping with concrete mechanisms: ahead-of-time (AOT) swap, stubs… | ||
| Retwork: Exploring Reader Network with COTS RFID Systems | Jia Liu, Xingyu Chen, Shigang Chen, Wei Wang, Dong Jiang, Lijun Chen | 18m |
Protocol identifies reader contention by piggybacking tag inventory flags to mimic a one-to-many write under EPC Gen2, then runs selective reads to mark… | ||
| Acclaim: Adaptive Memory Reclaim to Improve User Experience in Android Systems | Yu Liang, Jinheng Li, Rachata Ausavarungnirun, Riwei Pan, Liang Shi, Tei-Wei Kuo, Chun Jason Xue | 21m |
Acclaim implements a kernel reclaim policy with two concrete pieces: foreground aware eviction moves free pages from background apps to foreground ones, and… | ||
| SweynTooth: Unleashing Mayhem over Bluetooth Low Energy | Matheus E. Garbelini, Chundong Wang, Sudipta Chattopadhyay, Sun Sumei, Ernest Kurniawan | 25m |
An automated fuzzer controls the BLE link layer (using nRF52-based tester) and drives a state-machine model to send malformed, mistimed, and out-of-order… | ||
| Fine-Grained Isolation for Scalable, Dynamic, Multi-tenant Edge Clouds | Yuxin Ren, Guyue Liu, Vlad Nitu, Wenyuan Shao, Riley Kennedy, Gabriel Parmer, Timothy Wood, Alain Tchana | 21m |
EdgeOS implements a microkernel OS that builds featherweight processes and a copy-based memory movement accelerator to give per-client isolation while keeping… | ||
| Firefly: Untethered Multi-user VR for Commodity Mobile Devices | Xing Liu, Christina Vlachou, Feng Qian, Chendong Wang, Kyu-Han Kim | 25m |
System pre-renders panoramic 360° color+depth frames into tiled video (office map 137 GB, museum 99 GB), serves tiles with viewport-adaptive fetching plus… | ||
| nothing matches | ||