Jidong Zhai
Professor
(长聘教授、博士生导师)
PACMAN Lab
Institute of High Performance Computing
Department of Computer Science and Technology
Tsinghua University
zhaijidong AT tsinghua.edu.cn
Jidong Zhai (翟季冬) is a Full Professor in the Department of Computer Science and Technology of Tsinghua University. He was a Visiting Professor of Stanford University (2015-2016) and a Visiting Scholar of MSRA (Microsoft Research Asia) in 2013. He is the general secretary of ACM SIGHPC China.
His current research interests include parallel computing, compiler, programming language, and performance evaluation. He has published in prestigious conferences (such as SC, ICS, PPOPP, ASPLOS, MICRO, OSDI, ATC, and PACT) and top-tier journals (such as IEEE TC and IEEE TPDS). His research received 2021 Best Paper Award for IEEE TPDS, Best Paper Award at CLUSTER'21, Best Student Paper Award at ICS’21, Best Paper Honorable Mention Award at ICDCS’20, and Best Paper Finalist at SC’14. He has served as a PC member for a number of international conferences including SC, ICS, PPoPP, IPDPS, ICPP, Cluster, and PACT. He was a program co-chair of NPC 2018. He is currently on the editorial boards of IEEE Transactions on Computers (TC), IEEE Transactions on Parallel and Distributed Systems (TPDS), IEEE Transactions on Cloud Computing (TCC), Journal of Parallel and Distributed Computing (JPDC), and Journal of Parallel Computing (PARCO).
He is the advisor of Tsinghua Student Cluster Team. The team led by him has achieved 13 international champions in student supercomputing challenges at SC, ISC, and ASC. In 2015 and 2018, the team led by him swept all three champions. He has got a number of awards including NSFC Distinguished Young Scholar (2022), First Prize of CCF Natural Science Award (2021), CCF-IEEE CS Young Computer Scientist Award (2020), IEEE TPDS Award for Editorial Excellence (2019), NSFC Young Career Award (2017), and CCF outstanding doctoral dissertation Award (2010).
Latest
- MixQ
- MagPy
- WiseGraph
- A Survey on Performance Modeling and Prediction for Distributed DNN Training
- Compressed Data Direct Computing for Databases
- Editorial for the special issue on programming models and system software for High-Performance Computing (HPC) environments
- Efficient Inference for Pruned CNN Models on Mobile Devices With Holistic Sparsity Alignment
- F-TADOC: FPGA-Based Text Analytics Directly on Compression with HLS
- FastDecode: High-Throughput GPU-Efficient LLM Serving using Heterogeneous Pipelines
- FreshGNN: Reducing Memory Access via Stable Historical Embeddings for Graph Neural Network Training
- G-Learned Index: Enabling Efficient Learned Index on GPU
- Graph-Centric Performance Analysis for Large-Scale Parallel Applications
- MAGPY: Compiling Eager Mode DNN Programs by Monitoring Execution States
- MixQ: Taming Dynamic Outliers in Mixed-Precision Quantization by Online Prediction
- Optimal Kernel Orchestration for Tensor Programs with Korch
- Optimal Kernel Orchestration for Tensor Programs with Korch
- Optimizing I/O Performance Through Effective vCPU Scheduling Interference Management
- POSTER: Pattern-Aware Sparse Communication for Scalable Recommendation Model Training
- PUZZLE: Efficiently Aligning Large Language Models through Light-Weight Context Switch
- WiseGraph: Optimizing GNN with Joint Workload Partition of Graph and Operations
- GraphSet
- EinNet
- Vapro
- SmartMoE
- BladeDISC: Optimizing Dynamic Shape Machine Learning Workloads via Compiler Approach
- Cocktailer: Analyzing and Optimizing Dynamic Control Flow in Deep Learning
- CompressGraph: Efficient Parallel Graph Analytics with Rule-Based Compression
- Critique of \"A Parallel Framework for Constraint-Based Bayesian Network Learning via Markov Blanket Discovery\" by SCC Team From Tsinghua University
- EINNET: Optimizing Tensor Programs with Derivation-Based Transformations
- Enabling Efficient Random Access to Hierarchically Compressed Text Data on Diverse GPU Platforms
- GLM-130B: An Open Bilingual Pre-trained Model
- GraphSet: High Performance Graph Mining through Equivalent Set Transformations
- Joint Geometrical and Statistical Domain Adaptation for Cross-domain Code Vulnerability Detection
- Optimizing DNNs With Partially Equivalent Transformations and Automated Corrections
- PowerFusion: A Tensor Compiler with Explicit Data Movement Description and Instruction-level Graph IR
- ReFresh: Reducing Memory Access from Exploiting Stable Historical Embeddings for Graph Neural Network Training
- SmartMoE: Efficiently Training Sparsely-Activated Models through Combining Offline and Online Parallelization
- Special issue on new trends in high-performance computing: Software systems and applications
- Unified Programming Models for Heterogeneous High-Performance Computers
- Unveiling the Black Box of PLMs with Semantic Anchors: Towards Interpretable Neural Semantic Parsing
- APE
- FreeTensor
- BaGuaLu
- ZZ Crosstalk Suppression
- PerFlow
- FasterMoE
- AStitch: enabling a new multi-dimensional optimization space for memory-intensive ML training and inference on modern SIMT architectures
- BaGuaLu: targeting brain scale pretrained models with over 37 million cores
- CompressDB: Enabling Efficient Compressed Data Direct Processing for Various Databases
- Critique of \"MemXCT: Memory-Centric X-Ray CT Reconstruction With Massive Parallelization\" by SCC Team From Tsinghua University
- Detecting Performance Variance for Parallel Applications Without Source Code
- Efficiently emulating high-bitwidth computation with low-bitwidth hardware
- Exploring Data Analytics Without Decompression on Embedded GPU Systems
- Exploring Query Processing on CPU-GPU Integrated Edge Device
- FasterMoE: modeling and optimizing training of large-scale dynamic pre-trained models
- FreeTensor: a free-form DSL with holistic optimizations for irregular tensor programs
- GLM-130B: An Open Bilingual Pre-trained Model
- GraphQ IR: Unifying Semantic Parsing of Graph Query Language with Intermediate Representation
- GraphQ IR: Unifying the Semantic Parsing of Graph Query Languages with One Intermediate Representation
- Guest Editorial
- Guiding the PLMs with Semantic Anchors as Intermediate Supervision: Towards Interpretable Semantic Parsing
- Leveraging Code Snippets to Detect Variations in the Performance of HPC Systems
- Message from the High Performance Computing and Communications 2022 Program Chairs
- OLLIE: Derivation-based Tensor Program Optimizer
- Payment behavior prediction on shared parking lots with TR-GCN
- PerFlow: a domain specific framework for automatic performance analysis of parallel applications
- Periodic Weather-Aware LSTM With Event Mechanism for Parking Behavior Prediction
- POCLib: A High-Performance Framework for Enabling Near Orthogonal Processing on Compression
- Suppressing ZZ crosstalk of Quantum computers through pulse and scheduling co-optimization
- UniQ: A Unified Programming Model for Efficient Quantum Circuit Simulation
- Vapro: performance variance detection and diagnosis for production-run parallel applications
- Zoro: A robotic middleware combining high performance and high reliability
- Crosstalk Mitigation
- Elan
- CYPRESS
- ScalAna
- GNN Optimizations
- AIPerf
- FastMoE
- Graphpi
- HyQuas
- PET
- Spindle
- UniQ
- A Fast Lock for Explicit Message Passing Architectures
- Accelerating GPU Message Communication for Autonomous Navigation Systems
- AIPerf: Automated machine learning as an AI-HPC benchmark
- An Efficient Parallel Secure Machine Learning Framework on GPUs
- Automatic Irregularity-Aware Fine-Grained Workload Partitioning on Integrated Architectures
- Collaborative Heterogeneity-Aware OS Scheduler for Asymmetric Multicore Processors
- Critique of \"Planetary Normal Mode Computation: Parallel Algorithms, Performance, and Reproducibility\" by SCC Team From Tsinghua University
- FastMoE: A Fast Mixture-of-Expert Training System
- G-TADOC: Enabling Efficient GPU-Based Text Analytics without Decompression
- G-TADOC: Enabling Efficient GPU-Based Text Analytics without Decompression
- Guest Editorial
- HyQuas: hybrid partitioner based quantum circuit simulation system on GPU
- Mitigating Crosstalk in Quantum Computers through Commutativity-Based Instruction Reordering
- PET: Optimizing Tensor Programs with Partially Equivalent Transformations and Automated Corrections
- Preface
- TADOC: Text analytics directly on compression
- Understanding and bridging the gaps in current GNN performance optimizations
- AIPerf: Automated machine learning as an AI-HPC benchmark
- Edge-Stream: a Stream Processing Approach for Distributed Applications on a Hierarchical Edge-computing System
- Elan: Towards Generic and Efficient Elastic Training for Deep Learning
- Enabling Efficient Random Access to Hierarchically-Compressed Data
- GOPipe: A Granularity-Oblivious Programming Framework for Pipelined Stencil Executions on GPU
- GraphPi: high performance graph pattern matching through effective redundancy elimination
- GraphPi: High Performance Graph Pattern Matching through Effective Redundancy Elimination
- Identifying scalability bottlenecks for large-scale parallel programs with graph analysis
- Memory-Centric Communication Mechanism for Real-time Autonomous Navigation Applications
- Message Passing Optimization in Robot Operating System
- ParSecureML: An Efficient Parallel Secure Machine Learning Framework on GPUs
- Payment Behavior Prediction and Statistical Analysis for Shared Parking Lots
- PewLSTM: Periodic LSTM with Weather-Aware Gating Mechanism for Parking Behavior Prediction
- ScalAna: automating scaling loss detection with graph analysis
- ScalAna: Automating Scaling Loss Detection with Graph Analysis
- TADOC: Text Analytics Directly on Compression
- Automatic, Application-Aware I/O Forwarding Resource Allocation
- End-to-end I/O Monitoring on a Leading Supercomputer
- GOPipe: a granularity-oblivious programming framework for pipelined stencil executions on GPU
- Guest Editorial: Special Issue on Network and Parallel Computing for Emerging Architectures and Applications
- HiWayLib: A Software Framework for Enabling High Performance Communications for Heterogeneous Pipeline Computations
- Performance evaluation and analysis of sparse matrix and graph kernels on heterogeneous processors
- pLock: A Fast Lock for Architectures with Explicit Inter-core Message Passing
- Privacy Regulation Aware Process Mapping in Geo-Distributed Cloud Data Centers
- Spread-n-share: improving application performance and cluster throughput with resource-aware job placement
- Statistical Analysis and Prediction of Parking Behavior
- Student Cluster Competition 2018, Team Tsinghua University: Reproducing performance of multi-physics simulations of the Tsunamigenic 2004 Sumatra megathrust earthquake on the Intel Skylake Architecture
- A vision of post-exascale programming
- An adaptive breadth-first search algorithm on integrated architectures
- An Efficient In-Memory Checkpoint Method and its Practice on Fault-Tolerant HPL
- BitFlow: Exploiting Vector Parallelism for Binary Neural Networks on CPU
- CSE: Parallel Finite State Machines with Convergence Set Enumeration
- Efficient Document Analytics on Compressed Data: Method, Challenges, Algorithms, Insights
- Network and Parallel Computing - 15th IFIP WG 10.3 International Conference, NPC 2018, Muroran, Japan, November 29 - December 1, 2018, Proceedings
- Spindle: Informed Memory Access Monitoring
- Student cluster competition 2017, team Tsinghua University: Reproducing vectorization of the tersoff multi-body potential on the Intel Skylake and NVIDIA Volta architectures
- vSensor: leveraging fixed-workload snippets of programs for performance variance detection
- Zwift: A Programming Framework for High Performance Text Analytics on Compressed Data
- Algorithm-Directed Crash Consistence in Non-volatile Memory for HPC
- Algorithm-Directed Crash Consistence in Non-Volatile Memory for HPC
- Efficient process mapping in geo-distributed cloud data centers
- FinePar: irregularity-aware fine-grained workload partitioning on integrated architectures
- Scalable Graph Traversal on Sunway TaihuLight with Ten Million Cores
- Self-Checkpoint: An In-Memory Checkpoint Method Using Less Space and Its Practice on Fault-Tolerant HPL
- Understanding Co-Running Behaviors on Integrated CPU/GPU Architectures
- Versapipe: a versatile programming framework for pipelined computing on GPU
- A Fast Tridiagonal Solver for Intel MIC Architecture
- A survey of cloud resource management for complex engineering applications
- Building Semi-Elastic Virtual Clusters for Cost-Effective HPC Cloud Resource Provisioning
- Characterizing and optimizing TPC-C workloads on large-scale systems using SSD arrays
- Performance Prediction for Large-Scale Parallel Applications Using Representative Replay
- A Power-Conserving Online Scheduling Scheme for Video Streaming Services
- Automatic Cloud I/O Configurator for I/O Intensive Parallel Applications
- Optimizing seam carving on multi-GPU systems for real-time content-aware image resizing
- To Co-run, or Not to Co-run: A Performance Study on Integrated Architectures
- CYPRESS: Combining Static and Dynamic Analysis for Top-Down Communication Trace Compression
- Optimizing Seam Carving on multi-GPU systems for real-time image resizing
- ACIC: automatic cloud I/O configurator for HPC applications
- ACIC: automatic cloud I/O configurator for parallel applications
- Cost-effective cloud HPC resource provisioning by building semi-elastic virtual clusters
- Employing Checkpoint to Improve Job Scheduling in Large-Scale Systems
- Cloud versus in-house cluster: evaluating Amazon cluster compute instances for running MPI applications
- Efficiently Acquiring Communication Traces for Large-Scale Parallel Applications
- One optimized I/O configuration per HPC application: leveraging the configurability of cloud
- PHANTOM: predicting performance of parallel applications on large-scale parallel machines using a single node
- FACT: fast communication trace collection for parallel applications through program slicing
- LogGPO: An accurate communication model for performance prediction of MPI programs
- Process Mapping for MPI Collective Communications