Yupeng Gui
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Yupeng Gui has a total of 23 co-authors. Here are the top co-authors by number of shared publications:
- Yibo Fan
Fudan University
6 shared publications - Shichen Peng
School of Microelectronics, State Key Laboratory of Integrated Chips and Systems, Fudan University, Shanghai, China
5 shared publications - Ruiqi Tang
State Key Laboratory of Integrated Chips and Systems, Fudan University, Shanghai, China
4 shared publications - Xiaoyang Zeng
Fudan University
4 shared publications - Zhengyang Cao
Fudan University
4 shared publications - Minge Jing
Fudan University
3 shared publications
Latest Publications
- 11 November 2026
- Published by Elsevier in Integration
- Vol. 112
Bypass-bounded logarithmic response-ratio arbitration for SoC burst interfaces
- 11 November 2026
- Published by Elsevier in Integration
- Vol. 112
Bypass-bounded logarithmic response-ratio arbitration for SoC burst interfaces
- Open Access
- 22 June 2026
- Published by Association for Computing Machinery (ACM)
- p. 411-418
STP: Semantic-Triggered Prefetching for Event-Driven Workloads
- 22 June 2026
- Published by Association for Computing Machinery (ACM)
- p. 411-418
Open AccessSTP: Semantic-Triggered Prefetching for Event-Driven Workloads
- Open Access
- 18 August 2025
- Published by MDPI AG in Electronics
- Vol. 14 (16) , 3269
Memory-Efficient Feature Merging for Residual Connections with Layer-Centric Tile Fusion
Convolutional neural networks (CNNs) have achieved remarkable success in computer vision tasks, driving the rapid development of hardware accelerators. However, memory efficiency remains a key challenge, as conventional accelerators adopt layer-by-layer processing, leading to frequent external memory accesses (EMAs) of intermediate feature data, which increase energy consumption and latency. While layer fusion has been proposed to enhance inter-layer feature reuse, existing approaches typically rely on fixed data management tailored to specific architectures, introducing on-chip memory overhead and requiring trade-offs with EMAs. Moreover, prevalent residual connections further weaken fusion benefits due to diverse data reuse distances. To address these challenges, we propose layer-centric tile fusion, which integrates residual data loading with feature merging by leveraging receptive field relationships among feature tiles. A reuse distance-aware caching strategy is introduced to support flexible storage for various data types. We also develop a modeling framework to analyze the trade-off between on-chip memory usage and EMA-induced energy-delay product (EDP). Experimental results demonstrate that our method achieves 5.04–43.44% EDP reduction and 20.28–58.33% memory usage reduction compared to state-of-the-art designs on ResNet-18 and SRGAN.
- 18 August 2025
- Published by MDPI AG in Electronics
- Vol. 14 (16) , 3269
Open AccessMemory-Efficient Feature Merging for Residual Connections with Layer-Centric Tile Fusion
Convolutional neural networks (CNNs) have achieved remarkable success in computer vision tasks, driving the rapid development of hardware accelerators. However, memory efficiency remains a key challenge, as conventional accelerators adopt layer-by-layer processing, leading to frequent external memory accesses (EMAs) of intermediate feature data, which increase energy consumption and latency. While layer fusion has been proposed to enhance inter-layer feature reuse, existing approaches typically rely on fixed data management tailored to specific architectures, introducing on-chip memory overhead and requiring trade-offs with EMAs. Moreover, prevalent residual connections further weaken fusion benefits due to diverse data reuse distances. To address these challenges, we propose layer-centric tile fusion, which integrates residual data loading with feature merging by leveraging receptive field relationships among feature tiles. A reuse distance-aware caching strategy is introduced to support flexible storage for various data types. We also develop a modeling framework to analyze the trade-off between on-chip memory usage and EMA-induced energy-delay product (EDP). Experimental results demonstrate that our method achieves 5.04–43.44% EDP reduction and 20.28–58.33% memory usage reduction compared to state-of-the-art designs on ResNet-18 and SRGAN.
- 4 December 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Circuits and Systems for Video Technology
- Vol. 35 (5) , 4506-4520
A Tightly Coupled AI-ISP Vision Processor
To achieve high-quality and high-resolution image processing, this work presents a novel vision processor that facilitates deep learning-enhanced image processing pipelines. At the system level, by identifying that a divide-and-conquer approach is essential to synergize both classical image processing and image enhancement networks, we develop a tightly coupled system with strip-tile conversion dataflow to enable fine-grained low-latency data interactions between image signal processors (ISPs) and the deep learning accelerator (DLA). At the architecture level, we design a comprehensive set of 21 efficient image processing modules to construct classical ISP pipelines, a tile-based strip layer fusion DLA specifically optimized for networks, and a programmable pixel pool that seamlessly supports the data access patterns of the ISP and the DLA. At the software and hardware co-design level, we propose a comprehensive optimization framework to address the implementation overhead of networks while maintaining the image quality. Finally, evaluations of the AI-ISP vision processor demonstrate 53.95% external memory access reduction and 35.51% latency reduction, delivering superior image quality with minimal on-chip memory overhead. A throughput of up to 168.5 frames per second facilitates efficient processing of ultra-high definition (UHD) resolution images.
- 4 December 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Circuits and Systems for Video Technology
- Vol. 35 (5) , 4506-4520
A Tightly Coupled AI-ISP Vision Processor
To achieve high-quality and high-resolution image processing, this work presents a novel vision processor that facilitates deep learning-enhanced image processing pipelines. At the system level, by identifying that a divide-and-conquer approach is essential to synergize both classical image processing and image enhancement networks, we develop a tightly coupled system with strip-tile conversion dataflow to enable fine-grained low-latency data interactions between image signal processors (ISPs) and the deep learning accelerator (DLA). At the architecture level, we design a comprehensive set of 21 efficient image processing modules to construct classical ISP pipelines, a tile-based strip layer fusion DLA specifically optimized for networks, and a programmable pixel pool that seamlessly supports the data access patterns of the ISP and the DLA. At the software and hardware co-design level, we propose a comprehensive optimization framework to address the implementation overhead of networks while maintaining the image quality. Finally, evaluations of the AI-ISP vision processor demonstrate 53.95% external memory access reduction and 35.51% latency reduction, delivering superior image quality with minimal on-chip memory overhead. A throughput of up to 168.5 frames per second facilitates efficient processing of ultra-high definition (UHD) resolution images.