Hui Yu
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Hui Yu has a total of 56 co-authors. Here are the top co-authors by number of shared publications:
- Jin Zhao
Huazhong University of Science and Technology, Wuhan, China
13 shared publications - Yu Zhang
Huazhong University of Science and Technology
13 shared publications - Hai Jin
HuaZhong University of Science and Technology, Wuhan, Hubei, China
11 shared publications - Xiaofei Liao
HuaZhong University of Science and Technology, Wuhan, Hubei, China
9 shared publications - Ligang He
University of Warwick
7 shared publications - Haikun Liu
Huazhong University of Science and Technology
6 shared publications
Latest Publications
- Open Access
- 26 July 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-7
TDH-GNN: An Efficient Topology-driven Accelerator for Dynamic Heterogeneous GNN
Dynamic Heterogeneous Graph Neural Networks (DHGNNs) capture evolving structural and semantic information through metapath learning over dynamic heterogeneous graphs. However, existing solutions suffer from redundant metapath matching and excessive off-chip communication. To address these inefficiencies, we present TDH-GNN, a topology-driven accelerator for high-performance DHGNN inference. TDH-GNN introduces a hyperedge-centric incremental execution scheme, in which hyperedges encode dependencies among metapath instances to enable incremental metapath updates and eliminate redundant matching and computation. It also caches aggregated features of high-frequency edges to improve memory bandwidth efficiency. Implemented on a Xilinx Alveo U280 FPGA, TDH-GNN achieves average speedups of 143.5x and 42.9x, and average energy savings of 189.2x and 55.8x, over state-of-the-art DHGNN software systems on an Intel CPU and an A100 GPU, respectively. Compared with leading HGNN accelerators, TDH-GNN further delivers average speedups of 4.3x and 2.8x, together with energy savings of 5.9x and 3.7x.
- 26 July 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-7
Open AccessTDH-GNN: An Efficient Topology-driven Accelerator for Dynamic Heterogeneous GNN
Dynamic Heterogeneous Graph Neural Networks (DHGNNs) capture evolving structural and semantic information through metapath learning over dynamic heterogeneous graphs. However, existing solutions suffer from redundant metapath matching and excessive off-chip communication. To address these inefficiencies, we present TDH-GNN, a topology-driven accelerator for high-performance DHGNN inference. TDH-GNN introduces a hyperedge-centric incremental execution scheme, in which hyperedges encode dependencies among metapath instances to enable incremental metapath updates and eliminate redundant matching and computation. It also caches aggregated features of high-frequency edges to improve memory bandwidth efficiency. Implemented on a Xilinx Alveo U280 FPGA, TDH-GNN achieves average speedups of 143.5x and 42.9x, and average energy savings of 189.2x and 55.8x, over state-of-the-art DHGNN software systems on an Intel CPU and an A100 GPU, respectively. Compared with leading HGNN accelerators, TDH-GNN further delivers average speedups of 4.3x and 2.8x, together with energy savings of 5.9x and 3.7x.
- Open Access
- 26 July 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-7
A Locality-Aware Temporal Motif Mining Accelerator with Chunk-Based Search Tree Expansion
- 26 July 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-7
Open AccessA Locality-Aware Temporal Motif Mining Accelerator with Chunk-Based Search Tree Expansion
- Open Access
- 28 January 2026
- Published by Association for Computing Machinery (ACM)
- p. 591-604
DTMiner: A Data-Centric System for Efficient Temporal Motif Mining
- 28 January 2026
- Published by Association for Computing Machinery (ACM)
- p. 591-604
Open AccessDTMiner: A Data-Centric System for Efficient Temporal Motif Mining
- Open Access
- 15 November 2025
- Published by Association for Computing Machinery (ACM)
- p. 237-249
TaGNN: An Efficient Topology-aware Accelerator for High-performance Dynamic Graph Neural Network
- 15 November 2025
- Published by Association for Computing Machinery (ACM)
- p. 237-249
Open AccessTaGNN: An Efficient Topology-aware Accelerator for High-performance Dynamic Graph Neural Network
- Open Access
- 6 August 2025
- Published by Association for Computing Machinery (ACM)
- p. 230-246
TempGraph: An Efficient Chain-driven Temporal Graph Computing Framework on the GPU
- 6 August 2025
- Published by Association for Computing Machinery (ACM)
- p. 230-246
Open AccessTempGraph: An Efficient Chain-driven Temporal Graph Computing Framework on the GPU
- Open Access
- 30 March 2025
- Published by Association for Computing Machinery (ACM)
- p. 621-636
OHMiner: An Overlap-centric System for Efficient Hypergraph Pattern Mining
- 30 March 2025
- Published by Association for Computing Machinery (ACM)
- p. 621-636
Open AccessOHMiner: An Overlap-centric System for Efficient Hypergraph Pattern Mining
- Open Access
- 20 March 2025
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Architecture and Code Optimization
- Vol. 22 (1) , 1-26
An Efficient ReRAM-based Accelerator for Asynchronous Iterative Graph Processing
Graph processing has become a central concern for many real-world applications and is well-known for its low compute-to-communication ratios and poor data locality. By integrating computing logic into memory, resistive random access memory (ReRAM) tackles the demand for high memory bandwidth in graph processing. Despite the years’ research efforts, existing ReRAM-based graph processing approaches still face the challenges of redundant computation overhead . It is because the vertices of many subgraphs are ineffectively and repeatedly processed over the ReRAM crossbars for lots of iterations so as to update their states according to the vertices of other subgraphs regardless of the dependencies among the subgraphs. In this article, we propose ASGraph , a dependency-aware ReRAM-based graph processing accelerator that overcomes the aforementioned performance bottlenecks. Specifically, ASGraph dynamically constructs the subgraph based on the dependencies between vertices’ states and then detects constructed subgraph that owns high value (it is likely that it has accumulated many state propagations from its neighbors and is able to affect more other neighbors) to be preferentially processed. In this way, it makes the vertex states propagate along the dependencies between vertices as much as possible to reduce the redundant computation. Besides, ASGraph employs a hybrid processing scheme to accelerate the state propagations of the tightly connected subgraph, thereby minimizing the redundant computations. Experimental results show that ASGraph achieves 25.5× and 4.8× speedup and 70.8× and 2.2× energy saving on average compared with the state-of-the-art ReRAM-based graph processing accelerators, that is, GraphR and GaaS-X, respectively.
- 20 March 2025
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Architecture and Code Optimization
- Vol. 22 (1) , 1-26
Open AccessAn Efficient ReRAM-based Accelerator for Asynchronous Iterative Graph Processing
Graph processing has become a central concern for many real-world applications and is well-known for its low compute-to-communication ratios and poor data locality. By integrating computing logic into memory, resistive random access memory (ReRAM) tackles the demand for high memory bandwidth in graph processing. Despite the years’ research efforts, existing ReRAM-based graph processing approaches still face the challenges of redundant computation overhead . It is because the vertices of many subgraphs are ineffectively and repeatedly processed over the ReRAM crossbars for lots of iterations so as to update their states according to the vertices of other subgraphs regardless of the dependencies among the subgraphs. In this article, we propose ASGraph , a dependency-aware ReRAM-based graph processing accelerator that overcomes the aforementioned performance bottlenecks. Specifically, ASGraph dynamically constructs the subgraph based on the dependencies between vertices’ states and then detects constructed subgraph that owns high value (it is likely that it has accumulated many state propagations from its neighbors and is able to affect more other neighbors) to be preferentially processed. In this way, it makes the vertex states propagate along the dependencies between vertices as much as possible to reduce the redundant computation. Besides, ASGraph employs a hybrid processing scheme to accelerate the state propagations of the tightly connected subgraph, thereby minimizing the redundant computations. Experimental results show that ASGraph achieves 25.5× and 4.8× speedup and 70.8× and 2.2× energy saving on average compared with the state-of-the-art ReRAM-based graph processing accelerators, that is, GraphR and GaaS-X, respectively.
- Open Access
- 23 June 2024
- Published by Association for Computing Machinery (ACM)
- p. 1-6
CDA-GNN: A Chain-driven Accelerator for Efficient Asynchronous Graph Neural Network
Asynchronous Graph Neural Network (AGNN) has attracted much research attention because it enables faster convergence speed than the synchronous GNN. However, existing software/hardware solutions suffer from redundant computation overhead and excessive off-chip communications for AGNN due to irregular state propagations along the dependency chains between vertices. This paper proposes a chain-driven asynchronous accelerator, CDA-GNN, for efficient AGNN inference. Specifically, CDA-GNN proposes a chain-driven asynchronous execution approach into novel accelerator design to regularize the vertex state propagations for fewer redundant computations and off-chip communications and also designs a chain-aware data caching method to improve data locality for AGNN. We have implemented and evaluated CDA-GNN on a Xilinx Alveo U280 FPGA card. Compared with the cutting-edge software solutions (i.e., Dorylus and AMP) and hardware solutions (i.e., BlockGNN and FlowGNN), CDA-GNN improves the performance of AGNN inference by an average of 1,173x, 182.4x, 10.2x, and 7.9x and saves energy by 2,241x, 242.2x, 12.4x, and 8.9x, respectively.
- 23 June 2024
- Published by Association for Computing Machinery (ACM)
- p. 1-6
Open AccessCDA-GNN: A Chain-driven Accelerator for Efficient Asynchronous Graph Neural Network
Asynchronous Graph Neural Network (AGNN) has attracted much research attention because it enables faster convergence speed than the synchronous GNN. However, existing software/hardware solutions suffer from redundant computation overhead and excessive off-chip communications for AGNN due to irregular state propagations along the dependency chains between vertices. This paper proposes a chain-driven asynchronous accelerator, CDA-GNN, for efficient AGNN inference. Specifically, CDA-GNN proposes a chain-driven asynchronous execution approach into novel accelerator design to regularize the vertex state propagations for fewer redundant computations and off-chip communications and also designs a chain-aware data caching method to improve data locality for AGNN. We have implemented and evaluated CDA-GNN on a Xilinx Alveo U280 FPGA card. Compared with the cutting-edge software solutions (i.e., Dorylus and AMP) and hardware solutions (i.e., BlockGNN and FlowGNN), CDA-GNN improves the performance of AGNN inference by an average of 1,173x, 182.4x, 10.2x, and 7.9x and saves energy by 2,241x, 242.2x, 12.4x, and 8.9x, respectively.
- Open Access
- 23 June 2024
- Published by Association for Computing Machinery (ACM)
- p. 1-6
RTGA: A Redundancy-free Accelerator for High-Performance Temporal Graph Neural Network Inference
Temporal Graph Neural Network (TGNN) has attracted much research attention because it can capture the dynamic nature of complex networks. However, existing solutions suffer from redundant computation overhead and excessive off-chip communications for TGNN inference because they often rely on redundant graph sampling and repeatedly fetching the features and vertex memory. This paper proposes a redundancy-free accelerator, RTGA, for high-performance TGNN inference. Specifically, RTGA proposes a redundancy-aware execution approach with temporal tree into a novel accelerator design to effectively eliminate unnecessary data processing for fewer redundant computations and off-chip communications and also designs a temporal-aware data caching method to improve data locality for TGNN. We have implemented and evaluated RTGA on a Xilinx Alveo U280 FPGA card. Compared with cutting-edge software solutions (i.e., TGN and TGL) and hardware solutions (i.e., BlockGNN and FlowGNN), RTGA improves the performance of TGNN inference by an average of 473.2x, 87.4x, 8.2x, and 6.9x and saves energy by 542.8x, 102.2x, 9.4x, and 8.3x, respectively.
- 23 June 2024
- Published by Association for Computing Machinery (ACM)
- p. 1-6
Open AccessRTGA: A Redundancy-free Accelerator for High-Performance Temporal Graph Neural Network Inference
Temporal Graph Neural Network (TGNN) has attracted much research attention because it can capture the dynamic nature of complex networks. However, existing solutions suffer from redundant computation overhead and excessive off-chip communications for TGNN inference because they often rely on redundant graph sampling and repeatedly fetching the features and vertex memory. This paper proposes a redundancy-free accelerator, RTGA, for high-performance TGNN inference. Specifically, RTGA proposes a redundancy-aware execution approach with temporal tree into a novel accelerator design to effectively eliminate unnecessary data processing for fewer redundant computations and off-chip communications and also designs a temporal-aware data caching method to improve data locality for TGNN. We have implemented and evaluated RTGA on a Xilinx Alveo U280 FPGA card. Compared with cutting-edge software solutions (i.e., TGN and TGL) and hardware solutions (i.e., BlockGNN and FlowGNN), RTGA improves the performance of TGNN inference by an average of 473.2x, 87.4x, 8.2x, and 6.9x and saves energy by 542.8x, 102.2x, 9.4x, and 8.3x, respectively.
- Open Access
GIDN: A Lightweight Graph Inception Diffusion Network for High-efficient Link Prediction
In this paper, we propose a Graph Inception Diffusion Networks(GIDN) model. This model generalizes graph diffusion in different feature spaces, and uses the inception module to avoid the large amount of computations caused by complex network structures. We evaluate GIDN model on Open Graph Benchmark(OGB) datasets, reached an 11% higher performance than AGDN on ogbl-collab dataset.
Open AccessGIDN: A Lightweight Graph Inception Diffusion Network for High-efficient Link Prediction
In this paper, we propose a Graph Inception Diffusion Networks(GIDN) model. This model generalizes graph diffusion in different feature spaces, and uses the inception module to avoid the large amount of computations caused by complex network structures. We evaluate GIDN model on Open Graph Benchmark(OGB) datasets, reached an 11% higher performance than AGDN on ogbl-collab dataset.