会议日程
| 2022-05-21 (Day 1): | |
| 8:20 - 8:30 | 开幕 | KeynoteⅠ (Session Chair:华宇，华中科技大学) |
| 8:30 - 9:20 | Debugging Performance Issues in Modern Desktop Applications Junfeng Yang（Columbia University） |
| 9:20 - 10:10 | 高性能计算性能工程 谭光明（中国科学院计算技术研究所） |
| 10:10 - 11:00 | Acoustic sensing and applications Lili Qiu（Microsoft Research Asia and UT Austin） |
Session 1: Memory and Cache (Session Chair:杜子东，中国科学院计算研究所) |
| 11:00 - 11:20 | Segcache: a memory-efficient and scalable in-memory key-value cache for small objects 杨骏骋（卡内基梅隆大学）[演讲视频] |
| 11:20 - 11:40 | NVAlloc: rethinking heap metadata management in persistent memory allocators 党政（浙江大学）[演讲视频] |
| 11:40 - 12:00 | CacheSifter: Sifting Cache Files for Boosted Mobile Performance and Lifetime 梁宇（香港城市大学）[演讲视频] |
| 午休(12:00 - 13:00) | |
| Session 2: Cloud and DB (Session Chair:夏文，哈尔滨工业大学(深圳)) |
|
| 13:00 - 13:20 | Plugsched: A Safe and Efficient Live Update Approach for Cloud OS Scheduler 马腾（阿里巴巴）[演讲视频] |
| 13:20 - 13:40 | Fluid: Dataset Abstraction and Elastic Acceleration for Cloud-native Data-Intensive Applications 顾荣（南京大学）[演讲视频] |
| 13:40 - 14:00 | LOCAT: Low-Overhead Online Configuration Auto-Tuning of Spark SQL Applications 辛锦瀚（中国科学院深圳先进技术研究院）[演讲视频] |
| Session 3: Transactions (Session Chair:单一舟，华为技术有限公司) |
|
| 14:00 - 14:20 | Forerunner: Constraint-based Speculative Transaction Execution for Ethereum 郭众鑫（微软亚洲研究院）[演讲视频] |
| 14:20 - 14:40 | Aurogon: Taming Aborts in All Phases for Distributed In-Memory Transactions 姜天洋（清华大学）[演讲视频] |
| Session 4: Memory and Storage (Session Chair:糜泽羽，上海交通大学) |
|
| 14:40 - 15:00 | HTMFS: Strong Consistency Comes for Free with Hardware Transactional Memory in Persistent Memory File Systems 董明凯（上海交通大学）[演讲视频] |
| 15:00 - 15:20 | Clio: A Hardware-Software Co-Designed Disaggregated Memory System 单一舟（华为）[演讲视频] |
| 15:20 - 15:40 | Hardware-augmented page prefetching for disaggregated memory 李海锋（中国科学院计算技术研究所）[演讲视频] |
| 15:40 - 16:00 | Separating Data via Block Invalidation Time Inference for Write Amplification Reduction in Log-Structured Storage 王秋平（香港中文大学）[演讲视频] |
| 16:00 - 16:20 | MT2: Memory Bandwidth Regulation on Hybrid NVM/DRAM Platforms 董明凯（上海交通大学）[演讲视频] |
| Session 5: Bugs & Security (Session Chair:张源，复旦大学) |
|
| 16:20 - 16:40 | Understanding and Finding On-the-Fly Configuration Bugs 王腾（国防科技大学） |
| 16:40 - 17:00 | HyBP: Hybrid Isolation-Randomization Secure Branch Predictor 赵路坦（中国科学院信息工程研究所）[演讲视频] |
| Session 6: Concurrency (Session Chair:刘宇涛，华为技术有限公司) |
|
| 17:00 - 17:20 | Asymmetry-aware Scalable Locking 古金宇（上海交通大学）[演讲视频] |
| 17:20 - 17:40 | TileSpGEMM: A Tiled Algorithm for Parallel Sparse General Matrix-Matrix Multiplication on GPU 牛宇瑶（中国石油大学）[演讲视频] |
| 2022-05-22 (Day 2): | ||
| KeynoteⅡ (Session Chair:王肇国，上海交通大学) |
||
| 8:30 - 9:20 | 15 years of learning from mistakes in building system software Shan Lu（The University of Chicago）[演讲视频] |
|
| 9:20 - 10:10 | 处理器安全分支预测器 侯锐（中国科学院信息工程研究所）[演讲视频] |
|
| 10:10 - 11:00 | 基于DPU的存储创新 黄克骥（华为） |
|
| Session 7: Architecture for System (Ⅰ) (Session Chair:沈志荣，厦门大学) |
||
| 11:00 - 11:20 | Rolis: a software approach to efficiently replicating multi-core transactions 沈维海（纽约州立大学石溪分校）[演讲视频] |
|
| 11:20 - 11:40 | Romou: Rapidly Generate High-Performance Tensor Kernels for Mobile GPUs 梁任冬（加利福尼亚大学尔湾分校）[演讲视频] |
|
| 11:40 - 12:00 | Optimized MPI Collective Algorithms for Dragonfly Topology 丰光南（中山大学）[演讲视频] |
|
| 午休(12:00 - 13:00) | ||
| Session 8: Architecture for System (Ⅱ) (Session Chair:吴明瑜，上海交通大学) |
||
| 13:00 - 13:20 | Sanger: A Co-Design Framework for Enabling Sparse Attention using Reconfigurable Architecture 卢丽强（北京大学）[演讲视频] |
|
| 13:20 - 13:40 | 基于RISC-V的用户态虚拟机监控器DuVisor 糜泽羽 (上海交通大学)[演讲视频] |
|
| Session 9: Software Services (Session Chair:左鹏飞，华为技术有限公司) |
||
| 13:40 - 14:00 | FaaSFlow: Enable Efficient Workflow Execution for Function-as-a-Service 李子俊（上海交通大学）[演讲视频] |
|
| 14:00 - 14:20 | Web服务的自动服务器无感知卸载技术 吴明瑜（上海交通大学）[演讲视频] |
|
| 14:20 - 14:40 | Semantics Foundation for Cyber-Physical Systems Using Higher-Order UTP 徐雄（中国科学院软件研究所）[演讲视频] |
|
| 14:40 - 15:00 | INFless: A Native Serverless System for Low-Latency, High-Throughput Inference 杨亚南（天津大学）[演讲视频] |
|
| Session 10: System for ML (Session Chair:董明凯，上海交通大学) |
||
| 15:00 - 15:20 | NeutronStar: Distributed GNN Training with Hybrid Dependency Management 王千阁（东北大学）[演讲视频] |
|
| 15:20 - 15:40 | VELTAIR: towards high-performance multi-tenant deep learning services via adaptive compilation and scheduling 刘子汉 （上海交通大学）[演讲视频] |
|
| 15:40 - 16:00 | NASPipe: high performance and reproducible pipeline parallel supernet training via causal synchronous parallelism 赵世雄（香港大学）[演讲视频] |
|
| 16:00 - 16:20 | HET-GMP: A Graph-based System Approach to Scaling Large Embedding Model Training 苗旭鹏（北京大学）[演讲视频] |
|
| 16:20 - 16:40 | FasterMoE: Modeling and Optimizing Training of Large-Scale Dynamic Pre-Trained Models 何家傲（清华大学）[演讲视频] |
|
| Session 11: Compression (Session Chair:张权路，微软亚洲研究院) |
||
| 16:40 - 17:00 | CompressDB: Enabling Efficient Compressed Data Direct Processing for Various Databases 张峰（中国人民大学）[演讲视频] |
|
| 17:00 - 17:20 | TVStore: Automatically Bounding Time Series Storage via Time-Varying Compression 安彦哲（清华大学）[演讲视频] |
|
| 17:20 - 17:25 | 闭幕 | |