Xueyan Wang
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Xueyan Wang has a total of 101 co-authors. Here are the top co-authors by number of shared publications:
- Weisheng Zhao
National Key Laboratory of Spintronics, Hangzhou International Innovation Institute, Beihang University 1 , Hangzhou 311115
16 shared publications - Jianlei Yang
Beihang University
14 shared publications - Xiaotao Jia
School of Integrated Circuit Science and Engineering, Beihang University, Beijing, China
11 shared publications - Gang Qu
Department of Electrical and Computer Engineering & Institute for Systems Research, University of Maryland, College Park, MD, USA
7 shared publications - Song Bian
Beihang University, Beijing, China
4 shared publications - Yier Jin
University of Science and Technology of China
4 shared publications
Latest Publications
- Open Access
- 26 July 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-7
BiBNN: 88.6% Computation Reduction in Bayesian Neural Networks via Post-Training Bernoullization
- 26 July 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-7
Open AccessBiBNN: 88.6% Computation Reduction in Bayesian Neural Networks via Post-Training Bernoullization
- Open Access
- 26 July 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-7
SparseE: Unlocking Sparsities in Encrypted Sparse Matrix Multiplication via Hardware-Software Co-Design
- 26 July 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-7
Open AccessSparseE: Unlocking Sparsities in Encrypted Sparse Matrix Multiplication via Hardware-Software Co-Design
- Open Access
- 26 July 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-7
BSGCN: Taming Diverse Sparsities in GCNs via Adaptive Band-Segmentation and Tailored Caching
- 26 July 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-7
Open AccessBSGCN: Taming Diverse Sparsities in GCNs via Adaptive Band-Segmentation and Tailored Caching
- 14 May 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Circuits and Systems II: Express Briefs
- Vol. PP  (15497747) , 1
SACIM: A Synapse-inspired Asynchronous CIM with Discharging Feedback, Dynamic Delay and Parallelism Extension for CNN Inference
This brief presents a charge-domain analog compute-in-memory (ACIM) architecture, the Synapse-inspired Asynchronous CIM (SACIM), which addresses key challenges in generality, latency, and parallelism. The architecture includes three structures: a Non-Differential Decoupled (NDDC) bit-cell array for implementing negative signed weights; a Clock-free Asynchronous Controller (CFAC) to dynamically shorten latency; and an asynchronous Pulse-based Analog-to-Digital Converter (PADC) circuit with an NDDC coordination mechanism for parallelism extension. Fabricated in 40nm CMOS, the prototype achieves a peak energy efficiency of 655.78 TOPS/W and a throughput of 1180.4 GOPS, with 5.1 μW static power in a 99,280 μm2 area.
- 14 May 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Circuits and Systems II: Express Briefs
- Vol. PP  (15497747) , 1
SACIM: A Synapse-inspired Asynchronous CIM with Discharging Feedback, Dynamic Delay and Parallelism Extension for CNN Inference
This brief presents a charge-domain analog compute-in-memory (ACIM) architecture, the Synapse-inspired Asynchronous CIM (SACIM), which addresses key challenges in generality, latency, and parallelism. The architecture includes three structures: a Non-Differential Decoupled (NDDC) bit-cell array for implementing negative signed weights; a Clock-free Asynchronous Controller (CFAC) to dynamically shorten latency; and an asynchronous Pulse-based Analog-to-Digital Converter (PADC) circuit with an NDDC coordination mechanism for parallelism extension. Fabricated in 40nm CMOS, the prototype achieves a peak energy efficiency of 655.78 TOPS/W and a throughput of 1180.4 GOPS, with 5.1 μW static power in a 99,280 μm2 area.
- 29 September 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Circuits and Systems for Artificial Intelligence
- Vol. PP  (5) , 1-11
High-Performance Bayesian Neural Network Inference Accelerator Based on FPGA
Bayesian Neural Networks (BNNs) offer robust uncertainty estimation capabilities through probabilistic modeling, yet their prohibitively high computational complexity and resource consumption limit deployment in edge computing. In this paper, we propose an FPGA-based BNN inference accelerator that optimizes critical modules—including weight generation and Feed-Forward inference. By integrating pipelining techniques with distributed storage strategies, our design achieves a balanced trade-off between computational efficiency and resource utilization. Experimental results on the Xilinx ZYNQ7020 platform demonstrate that, at a 100 MHz clock frequency, the accelerator achieves a single-inference latency of 0.05 seconds, achieving speedups of 254.8× and 6.2× over CPU and GPU platforms, respectively, while offering 28.4× higher energy efficiency than GPUs and maintaining a recognition accuracy of 98.4%.
- 29 September 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Circuits and Systems for Artificial Intelligence
- Vol. PP  (5) , 1-11
High-Performance Bayesian Neural Network Inference Accelerator Based on FPGA
Bayesian Neural Networks (BNNs) offer robust uncertainty estimation capabilities through probabilistic modeling, yet their prohibitively high computational complexity and resource consumption limit deployment in edge computing. In this paper, we propose an FPGA-based BNN inference accelerator that optimizes critical modules—including weight generation and Feed-Forward inference. By integrating pipelining techniques with distributed storage strategies, our design achieves a balanced trade-off between computational efficiency and resource utilization. Experimental results on the Xilinx ZYNQ7020 platform demonstrate that, at a 100 MHz clock frequency, the accelerator achieves a single-inference latency of 0.05 seconds, achieving speedups of 254.8× and 6.2× over CPU and GPU platforms, respectively, while offering 28.4× higher energy efficiency than GPUs and maintaining a recognition accuracy of 98.4%.
- 16 June 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems
- Vol. 45  (1) , 246-259
Efficient SRAM-PIM Co-Design by Joint Exploration of Value-Level and Bit-Level Sparsity
Processing-in-memory (PIM) architectures mitigate the Von Neumann bottleneck by integrating computation units into memory arrays. Among PIM architectures, digital SRAMPIM has become a prominent approach, directly integrating digital logic within the SRAM array. However, the rigid crossbar architecture and full array activation pose challenges in efficiently utilizing value-level sparsity. Moreover, neural network models exhibit a high proportion of zero bits within non-zero values, which remain underutilized due to architectural constraints. To overcome these limitations, we present Dyadic Block PIM (DB-PIM), a groundbreaking algorithm-architecture co-design framework to harness both value-level and bit-level sparsity. At the algorithm level, our hybrid-grained pruning technique, combined with a novel sparsity pattern, enables effective sparsity management. Architecturally, DB-PIM incorporates a sparse network and customized digital SRAM-PIM macros, including input pre-processing unit (IPU), dyadic block multiply units (DBMUs), and Canonical Signed Digit (CSD)-based adder trees. It circumvents structured zero values in weights and bypasses unstructured zero bits within non-zero weights and block-wise all-zero bit columns in input features. As a result, the DBPIM framework skips a majority of unnecessary computations, thereby driving significant gains in computational efficiency. Experimental results demonstrate that our DB-PIM framework achieves up to 8.01× speedup and 85.28% energy savings, significantly boosting computational efficiency in digital SRAMPIM systems.
- 16 June 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems
- Vol. 45  (1) , 246-259
Efficient SRAM-PIM Co-Design by Joint Exploration of Value-Level and Bit-Level Sparsity
Processing-in-memory (PIM) architectures mitigate the Von Neumann bottleneck by integrating computation units into memory arrays. Among PIM architectures, digital SRAMPIM has become a prominent approach, directly integrating digital logic within the SRAM array. However, the rigid crossbar architecture and full array activation pose challenges in efficiently utilizing value-level sparsity. Moreover, neural network models exhibit a high proportion of zero bits within non-zero values, which remain underutilized due to architectural constraints. To overcome these limitations, we present Dyadic Block PIM (DB-PIM), a groundbreaking algorithm-architecture co-design framework to harness both value-level and bit-level sparsity. At the algorithm level, our hybrid-grained pruning technique, combined with a novel sparsity pattern, enables effective sparsity management. Architecturally, DB-PIM incorporates a sparse network and customized digital SRAM-PIM macros, including input pre-processing unit (IPU), dyadic block multiply units (DBMUs), and Canonical Signed Digit (CSD)-based adder trees. It circumvents structured zero values in weights and bypasses unstructured zero bits within non-zero weights and block-wise all-zero bit columns in input features. As a result, the DBPIM framework skips a majority of unnecessary computations, thereby driving significant gains in computational efficiency. Experimental results demonstrate that our DB-PIM framework achieves up to 8.01× speedup and 85.28% energy savings, significantly boosting computational efficiency in digital SRAMPIM systems.
- Open Access
- 23 June 2024
- Published by Association for Computing Machinery (ACM)
- p. 1-6
Towards Efficient SRAM-PIM Architecture Design by Exploiting Unstructured Bit-Level Sparsity
- 23 June 2024
- Published by Association for Computing Machinery (ACM)
- p. 1-6
Open AccessTowards Efficient SRAM-PIM Architecture Design by Exploiting Unstructured Bit-Level Sparsity