会议日程
| 2022-12-17 (Day 1): | |
| 8:20 - 8:30 | 开幕:大会主席、程序委员会主席 | KeynoteⅠ (Session Chair: 毛波, 厦门大学) |
| 8:30 - 9:10 | 分形冯诺依曼体系结构 陈云霁（中国科学院计算技术研究所） |
| 9:10 - 9:50 | 面向人机物融合的泛在操作系统：机遇与挑战 郭耀（北京大学） |
| 9:50 - 10:30 | 操作系统“安全税”与系统软硬件协同方法 夏虞斌（上海交通大学） |
| 10:30 - 11:10 | 大规模分布式文件系统的技术挑战与研究进展 曾令仿（之江实验室） |
ChinaSys 新星奖和优秀博士论文奖颁奖（ChinaSys 主席：包云岗） |
| Session 1: Architecture For System (Session Chair: 赵永威, 中国科学院计算技术研究所) |
|
| 11:25 - 11:42 | A Bitflow Architecture for Arbitrary Precision Computing 郝一帆 (中国科学院计算技术研究所, 寒武纪) |
| 11:42 - 11:59 | Towards Developing High Performance RISC-V Processors Using Agile Methodology 徐易难 (中国科学院计算技术研究所, 中国科学院大学) |
| 11:59 - 12:16 | FpgaNIC: An FPGA-based Versatile 100Gb SmartNIC for GPUs Allowing Flexible Architecture Exploration 黄鸿靖 (浙江大学) |
| 午休(12:16 - 13:30) | |
| Session 2: Machine Learning I (Session Chair: 李诗逸, 哈尔滨工业大学(深圳)) |
|
| 13:30 - 13:47 | SparTA: Deep-Learning Model Sparsity via Tensor-with-Sparsity-Attribute 郑宁馨 (微软研究院) |
| 13:47 - 14:04 | Primo: Practical Learning-Augmented Systems with Interpretable Models 胡擎昊 (南洋理工大学) |
| 14:04 - 14:21 | ROLLER: Fast and Efficient Tensor Compilation for Deep Learning 朱虹宇 (多伦多大学, 微软研究院) |
| 14:21 - 14:38 | CoGNN: Efficient Scheduling for Concurrent GNN Training on GPUs 孙庆骁 (北京航空航天大学) |
| 14:38 - 14:55 | SOTER: Guarding black-box Inference for General Neural Networks at the Edge 沈天翔 (香港大学) |
| Session 3: Security & Privacy I (Session Chair: 张咪, 中国科学院计算技术研究所) |
|
| 15:00 - 15:17 | Efficient Distributed Secure Memory with Migratable Merkle Tree 冯二虎 (上海交通大学) |
| 15:17 - 15:34 | TensorFHE: Achieving Practical Computation on Encrypted Data Using GPGPU 范胜玉 (中国科学院信息工程研究所) |
| 15:34 - 15:51 | SMILE: Secure Memory Introspection for Live Enclave 周雷 (南方科技大学) |
| Session 4: NVM and Memory (Session Chair: 张杰, 北京大学) |
|
| 15:56 - 16:13 | RETROFIT: Realizing Extreme Endurance Through Fault-aware Wear Leveling and Improved Tolerance 张江伟 (清华大学) |
| 16:13 - 16:30 | ESD: An ECC-assisted and Selective Deduplication for Encrypted NonVolatile Main Memory 杜春锋 (厦门大学) |
| 16:30 - 16:47 | Hardware-Supported Remote Persistence for Distributed Persistent Memory 段卓辉 (华中科技大学) |
| 16:47 - 17:04 | FlatFS: Flatten Hierarchical File System Namespace on Non-volatile Memories 蔡淼 (河海大学, 南京大学) |
| 17:04 - 17:21 | HoPP: Hardware-Software Co-Designed Page Prefetching for Disaggregated Memory 李海锋 (中国科学院计算技术研究所) |
| Session 5: IoT/Mobile Systems (Session Chair: 李乔, 厦门大学) |
|
| 17:26 - 17:43 | ROG: A High Performance and Robust Distributed Training System for Robotic IoT 管修贤 (香港大学) |
| 17:43 - 18:00 | Ice: 基于“内存-进程”协同管理的智能终端冻结系统 李昌龙 (华东师范大学, 香港城市大学) |
| 18:00 - 18:17 | Using Psychophysics to Guide Power Adaptation for Input Methods on Mobile Architectures 李雪亮 (深圳大学) |
| 2022-12-18 (Day 2): | ||
| KeynoteⅡ (Session Chair: 王晨曦, 中国科学院计算技术研究所) |
||
| 8:30 - 9:10 | 泛在操作系统：系统之系统的认识与思考 李宣东 (南京大学) |
|
| 9:10 - 9:50 | Constructing Benchmark Suites from Real Cloud Applications: Practice and Challenges 喻之斌（华为云竖亥实验室） |
|
| 9:50 - 10:30 | 面向超大规模数据的存算协同 张一鸣 (厦门大学) |
|
| Session 6: Machine Learning II (Session Chair: 赵家程, 中国科学院计算技术研究所) |
||
| 10:35 - 10:52 | Microsecond-scale Preemption for Concurrent GPU-accelerated DNN Inferences 韩明聪 (上海交通大学, 上海人工智能实验室) |
|
| 10:52 - 11:09 | Coop: Unleashing Potential of Your GPU with Co-optimization of Tensor Rematerialization and Allocation 张建浩 (一流科技) |
|
| 11:09 - 11:26 | OAMOS: Enabling Automatic Mapping for Tensor Computations On Spatial Accelerators with Hardware Abstraction 郑思泽 (北京大学) |
|
| 11:26 - 11:43 | Galvatron: Efficient Transformer Training over Multiple GPUs Using Automatic Parallelism 苗旭鹏 (北京大学, 卡内基梅隆大学) |
|
| 11:43 - 12:00 | MPress: Democratizing Billion-Scale Model Training on Multi-GPU Servers via Memory-Saving Inter-Operator Parallelism 周泉 (中国科学技术大学) |
|
| 午休(12:00 - 13:30) | ||
| Announcement Session | ||
| 13:30 - 13:40 | ChinaSys 学术开源创新平台介绍 王肇国（上海交通大学） |
|
| Session 7: Security & Privacy II (Session Chair: 左志强，南京大学) |
||
| 13:45 - 14:02 | HyperEnclave: An Open and Cross-platform Trusted Execution Environment 贾越凯 (清华大学) |
|
| 14:02 - 14:19 | Cronus: Fault-isolated, Secure and High-performance Heterogeneous Computing for Trusted Execution Environment 江健宇 (香港大学) |
|
| 14:19 - 14:36 | StrongBox: A GPU TEE on Arm Endpoints 邓韵杰 (南方科技大学) |
|
| Session 8: Concurrency and Parallelism Strategy (Session Chair: 吴明瑜, 上海交通大学) |
||
| 14:41 - 14:58 | Boosting Performance and QoS for Concurrent GPU B+trees by Combining-based Synchronization 赵传磊 (复旦大学) |
|
| 14:58 - 15:15 | Isolation Mechanisms for High-Speed Packet-Processing Pipelines 杨翔瑞 (国防科技大学) |
|
| 15:15 - 15:32 | Parallel Training of Pre-trained Models via Chunk-based Dynamic Memory Management 方佳瑞 (腾讯, HPCAI-Tech) |
|
| 15:32 - 15:49 | Efficient Supernet Training Using Path Parallelism 徐颖(中国科学院计算技术研究所) |
|
| Session 9: Microservice & Serverless (Session Chair: 王卅, 中国科学院计算技术研究所) |
||
| 15:54 - 16:11 | Erms: Efficient Resource Management for Shared Microservices with SLA Guarantees 罗树添(中国科学院深圳先进技术研究院, 中国科学院大学, 澳门大学) |
|
| 16:11 - 16:28 | No Provisioned Concurrency: Fast RDMA-codesigned Remote Fork for Serverless Computing 魏星达 (上海交通大学, 上海人工智能实验室) |
|
| Session 10: Potpourri (Session Chair: 蒋炎岩，南京大学) |
||
| 16:33 - 16:50 | CompressStreamDB: Fine-Grained Adaptive Stream Processing without Decompression 张宇 (中国人民大学) |
|
| 16:50 - 17:07 | HyQSAT: A Hybrid Approach for 3-SAT Problems by Integrating Quantum Annealer with CDCL 谭思危 (浙江大学) |
|
| 17:07 - 17:24 | H基于数据分布特征的稀疏矩阵向量乘优化研究 高建花 (北京理工大学) |
|
| 17:24 - 17:41 | LightTraffic: On Optimizing CPU-GPU Data Traffic for Efficient Large-scale Random Walks 邢益鹏 (中国科学技术大学) |
|
| 17:41 - 17:46 | 闭幕 | |