Wenjie Fan
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Wenjie Fan has a total of 22 co-authors. Here are the top co-authors by number of shared publications:
- Li Li
Nanjing University
5 shared publications - Yuxiang Fu
Nanjing University
5 shared publications - Siyue Li
School of Integrated Circuits and the School of Electronic Science and Engineering, Nanjing University, Qixia, Nanjing, Jiangsu, China
3 shared publications Chenyang Dai
School of Integrated Circuits and the School of Electronic Science and Engineering, Nanjing University, Qixia, Nanjing, Jiangsu, China
2 shared publications- Heng Zhang
Nanjing University
2 shared publications - Jinlun Ji
School of Integrated Circuits, and School of Electronic Science and Engineering, Nanjing University, Nanjing, Jiangsu, China
2 shared publications
Latest Publications
- 6 October 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Circuits and Systems II: Express Briefs
- Vol. 72 (12) , 2002-2006
NMIX: NoC-Aware Mixed Precision Quantization for Energy-Efficient ML Accelerator
Energy efficiency remains a critical challenge in network-on-chip (NoC) based machine learning (ML) accelerators due to high computation and communication costs. Existing NoC optimization techniques often overlook the intrinsic co-optimization between NoC characteristics and the ML model design. We propose NoC-aware mixed precision quantization (NMIX), a two-phase algorithm that jointly minimizes computation and routing energy while preserving model accuracy. The first phase performs a coarse search to identify accuracy-sensitive quantization boundaries, followed by fine-grained precision assignment based on sensitivity and NoC energy profiles. Experiments on image classification tasks with VGG and ResNet demonstrate energy savings of up to 91.5% over FP32 and 57.5% over uniform 8-bit quantization, confirming the scalability and effectiveness of NMIX for energy-efficient ML accelerators.
- 6 October 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Circuits and Systems II: Express Briefs
- Vol. 72 (12) , 2002-2006
NMIX: NoC-Aware Mixed Precision Quantization for Energy-Efficient ML Accelerator
Energy efficiency remains a critical challenge in network-on-chip (NoC) based machine learning (ML) accelerators due to high computation and communication costs. Existing NoC optimization techniques often overlook the intrinsic co-optimization between NoC characteristics and the ML model design. We propose NoC-aware mixed precision quantization (NMIX), a two-phase algorithm that jointly minimizes computation and routing energy while preserving model accuracy. The first phase performs a coarse search to identify accuracy-sensitive quantization boundaries, followed by fine-grained precision assignment based on sensitivity and NoC energy profiles. Experiments on image classification tasks with VGG and ResNet demonstrate energy savings of up to 91.5% over FP32 and 57.5% over uniform 8-bit quantization, confirming the scalability and effectiveness of NMIX for energy-efficient ML accelerators.
- 1 January 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Circuits and Systems I: Regular Papers
- Vol. 72 (9) , 4710-4722
FAS-NoC: A Real-Time Fused Approximation Scheme Coordinating Communication and Computation for NoC-Based NN Accelerators
Network-on-Chip (NoC) is a scalable on-chip communication architecture widely used in neural network accelerators. However, data-intensive applications like machine learning place significant demands on the NoC’s communication and computation, and often have a degree of resilience to data noise, which allows to use approximation techniques to reduce execution time and energy consumption for both computation and communication, under the constraints of acceptable quality loss. Traditional approximate NoCs do not consider the data distribution characteristics of the neural networks, resulting in a lower approximate rate. Moreover, these schemes do not take the synergistic optimization of computation and communication, which limits reductions in execution time. In this paper, we propose a Fused Approximation Scheme of NoC (FAS-NoC) that incorporates the characteristics of data distribution in neural networks. FAS-NoC includes an approximate compression and recovery scheme based on data hierarchy, and uses a congestion-aware scheme to adjust the approximate rate of the node. Additionally, by leveraging the characteristics of recovered data after approximate communication, the scheme optimizes the design of computing units within the computing array. FAS-NoC collaboratively optimizes approximate communication and computation, organically integrating the two aspects. Compared with the state-of-the-art approximate framework ACDC (ACDC_ABDTR and ACDC_APPROX), the execution time of FAS-NoC is reduced by 48.94% and 47.72%, respectively. The experimental results show that the additional area overhead of the FAS-NoC only accounts for 0.81% of the original node, the additional power consumption overhead only accounts for 0.75% of the original node.
- 1 January 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Circuits and Systems I: Regular Papers
- Vol. 72 (9) , 4710-4722
FAS-NoC: A Real-Time Fused Approximation Scheme Coordinating Communication and Computation for NoC-Based NN Accelerators
Network-on-Chip (NoC) is a scalable on-chip communication architecture widely used in neural network accelerators. However, data-intensive applications like machine learning place significant demands on the NoC’s communication and computation, and often have a degree of resilience to data noise, which allows to use approximation techniques to reduce execution time and energy consumption for both computation and communication, under the constraints of acceptable quality loss. Traditional approximate NoCs do not consider the data distribution characteristics of the neural networks, resulting in a lower approximate rate. Moreover, these schemes do not take the synergistic optimization of computation and communication, which limits reductions in execution time. In this paper, we propose a Fused Approximation Scheme of NoC (FAS-NoC) that incorporates the characteristics of data distribution in neural networks. FAS-NoC includes an approximate compression and recovery scheme based on data hierarchy, and uses a congestion-aware scheme to adjust the approximate rate of the node. Additionally, by leveraging the characteristics of recovered data after approximate communication, the scheme optimizes the design of computing units within the computing array. FAS-NoC collaboratively optimizes approximate communication and computation, organically integrating the two aspects. Compared with the state-of-the-art approximate framework ACDC (ACDC_ABDTR and ACDC_APPROX), the execution time of FAS-NoC is reduced by 48.94% and 47.72%, respectively. The experimental results show that the additional area overhead of the FAS-NoC only accounts for 0.81% of the original node, the additional power consumption overhead only accounts for 0.75% of the original node.
- Open Access
- 12 June 2024
- Published by Association for Computing Machinery (ACM)
- p. 364-369
TTNNM: Thermal- and Traffic-Aware Neural Network Mapping on 3D-NoC-based Accelerator
3D Network on Chips (3D-NoCs) have ample on-chip wiring resources and high bandwidth, yet face numerous hotspots and higher temperature gradients due to increased integration and power density. This could lead to device failure, impacting system stability. Our paper introduces a thermal- and traffic-aware mapping method for 3D-NoC-based neural network accelerators. Firstly, based on the average load of different neural network layer, we determine their mapping sequences and suitable dies. Secondly, to minimize delay and alleviate hotspot temperatures, we allocate groups to appropriate nodes. Compared with previous works, TTNNM reduces the average temperature by 3.0°C, 2.2°C, 2.4°C, temperature variance by 58.4%, 64.8%, 73.0%, maximum temperature by 9.3°C, 7.9°C, 12.0°C, and packet latency by 31.7%, 17.2%, 25.1%.
- 12 June 2024
- Published by Association for Computing Machinery (ACM)
- p. 364-369
Open AccessTTNNM: Thermal- and Traffic-Aware Neural Network Mapping on 3D-NoC-based Accelerator
3D Network on Chips (3D-NoCs) have ample on-chip wiring resources and high bandwidth, yet face numerous hotspots and higher temperature gradients due to increased integration and power density. This could lead to device failure, impacting system stability. Our paper introduces a thermal- and traffic-aware mapping method for 3D-NoC-based neural network accelerators. Firstly, based on the average load of different neural network layer, we determine their mapping sequences and suitable dies. Secondly, to minimize delay and alleviate hotspot temperatures, we allocate groups to appropriate nodes. Compared with previous works, TTNNM reduces the average temperature by 3.0°C, 2.2°C, 2.4°C, temperature variance by 58.4%, 64.8%, 73.0%, maximum temperature by 9.3°C, 7.9°C, 12.0°C, and packet latency by 31.7%, 17.2%, 25.1%.
- 30 April 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Circuits and Systems II: Express Briefs
- Vol. 71 (10) , 4521-4525
Communication Synchronization-Aware Arbitration Policy in NoC-Based DNN Accelerators
In NoC-based neural network accelerators, many-to-one and many-to-many are prevalent traffic patterns. In these traffic patterns, there exists a need for communication synchronization between Processing Elements (PEs) of adjacent layers to optimize latency. The last received packet will determine the end time of a layer’s computation. The Communication Synchronization-aware Arbitration Policy (CSAP) is proposed in this paper to handle this problem, which uses a negative feedback mechanism to regulate the packet sending rate of each source node. Compared with the local-age-based policy, CSAP decreases the execution by 4.69% ~ 12.55% across neural networks of different scales. The proposed policy only incurs 1.06% additional hardware overhead in the router compared with the local-age-based policy.
- 30 April 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Circuits and Systems II: Express Briefs
- Vol. 71 (10) , 4521-4525
Communication Synchronization-Aware Arbitration Policy in NoC-Based DNN Accelerators
In NoC-based neural network accelerators, many-to-one and many-to-many are prevalent traffic patterns. In these traffic patterns, there exists a need for communication synchronization between Processing Elements (PEs) of adjacent layers to optimize latency. The last received packet will determine the end time of a layer’s computation. The Communication Synchronization-aware Arbitration Policy (CSAP) is proposed in this paper to handle this problem, which uses a negative feedback mechanism to regulate the packet sending rate of each source node. Compared with the local-age-based policy, CSAP decreases the execution by 4.69% ~ 12.55% across neural networks of different scales. The proposed policy only incurs 1.06% additional hardware overhead in the router compared with the local-age-based policy.
- 12 February 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Circuits and Systems I: Regular Papers
- Vol. 71 (4) , 1863-1875
HAS-RL: A Hierarchical Approximate Scheme Optimized With Reinforcement Learning for NoC-Based NN Accelerators
Network-on-Chip (NoC) is a scalable on-chip communication architecture for the NN accelerator, but with the increase in the number of nodes, the communication delay becomes higher. Applications such as machine learning have a certain resilience to noisy/erroneous transmitted data. Therefore, approximate communication becomes a promising solution to improving performance by reducing traffic loads under the constraint of the acceptable maximum accuracy loss of neural networks. It is a key issue to balance the result quality and the communication delay for approximate NoC systems. The traditional approximate NoC only considers the node-to-node approximation-based dynamic traffic regulation. However, the dynamically changing traffic patterns across different nodes, different times, and different applications lead to a huge search space, which makes it hard to explore an optimal global approximation solution. In this paper, we propose a quality model for different neural networks, which presents the relationship between the quality loss and the data approximate rate. Then, a hierarchical approximate scheme optimized with reinforcement learning (HAS-RL) is proposed and we reduce the complexity of the HAS-RL by reducing the state space and action space, which will reduce the resource overhead as well. After that, we embed a global approximate controller in the NoC system, in which we deploy a policy network trained with the offline reinforcement learning algorithm to adjust the data approximate rates of each node at run time. Compared with the state-of-the-art method, the proposed scheme reduces the average network delay by 13.5% while their accuracies are similar. The proposed HAS-RL only causes an additional area overhead of 1.24% and power consumption of 0.77% compared with the traditional router design.
- 12 February 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Circuits and Systems I: Regular Papers
- Vol. 71 (4) , 1863-1875
HAS-RL: A Hierarchical Approximate Scheme Optimized With Reinforcement Learning for NoC-Based NN Accelerators
Network-on-Chip (NoC) is a scalable on-chip communication architecture for the NN accelerator, but with the increase in the number of nodes, the communication delay becomes higher. Applications such as machine learning have a certain resilience to noisy/erroneous transmitted data. Therefore, approximate communication becomes a promising solution to improving performance by reducing traffic loads under the constraint of the acceptable maximum accuracy loss of neural networks. It is a key issue to balance the result quality and the communication delay for approximate NoC systems. The traditional approximate NoC only considers the node-to-node approximation-based dynamic traffic regulation. However, the dynamically changing traffic patterns across different nodes, different times, and different applications lead to a huge search space, which makes it hard to explore an optimal global approximation solution. In this paper, we propose a quality model for different neural networks, which presents the relationship between the quality loss and the data approximate rate. Then, a hierarchical approximate scheme optimized with reinforcement learning (HAS-RL) is proposed and we reduce the complexity of the HAS-RL by reducing the state space and action space, which will reduce the resource overhead as well. After that, we embed a global approximate controller in the NoC system, in which we deploy a policy network trained with the offline reinforcement learning algorithm to adjust the data approximate rates of each node at run time. Compared with the state-of-the-art method, the proposed scheme reduces the average network delay by 13.5% while their accuracies are similar. The proposed HAS-RL only causes an additional area overhead of 1.24% and power consumption of 0.77% compared with the traditional router design.