Research and Practice on Co-Optimization of GPU and FPGA in Real-Time Hardware Generation
DOI:
https://doi.org/10.71222/maab5t81Keywords:
GPU, FPGA, collaborative optimization, real-time hardware generation, deep learningAbstract
With the increase of real-time computing requirements, GPU and FPGA collaborative optimization has become a key technology to improve hardware performance. In this paper, the collaborative computing model and architecture of GPU and FPGA are discussed and applied in image processing, signal processing, deep learning and other fields. By optimizing computing models, algorithms, data transmission and parallel computing, the computing speed and resource utilization are significantly improved. At the same time, optimization strategies such as resource scheduling, load balancing, and power consumption management are also proposed. Through experimental verification, it shows the wide application prospect and practical effect of GPU and FPGA cooperative optimization in real-time hardware generation.
References
1. Y. Xie, Z. Zhong, B. Li, Y. Xie, L. Chen, H. Chen, et al., "An ARM-FPGA hybrid acceleration and fault tolerant technique for phase factor calculation in spaceborne synthetic aperture radar imaging," IEEE J. Sel. Top. Appl. Earth Observ. Remote Sens., vol. 17, pp. 5059–5072, 2024, doi: 10.1109/JSTARS.2024.3365464.
2. M. Liu, Y. Wang, and S. Li, "A field programmable gate array placement methodology for netlist-level circuits with GPU acceleration," Electronics, vol. 13, no. 1, p. 37, 2023, doi: 10.3390/electronics13010037.
3. R. Moghanni and A. Hakkaki-Fard, "Optimizing vertical ground heat exchanger modelling through GPU-accelerated com-putation strategies," Renew. Energy, vol. 221, p. 119790, 2024, doi: 10.1016/j.renene.2023.119790.
4. S. Liu, W. Feng, J. Zhao, Z. Zhao, X. Liu, R. Liu, et al., "Collaborative optimization model of blast furnace raw materials and operating parameters based on intelligent calculation," ISIJ Int., vol. 64, no. 8, pp. 1229–1239, 2024, doi: 10.2355/isijinternational.ISIJINT-2023-450
Downloads
Published
Issue
Section
License
Copyright (c) 2025 Huijie Pan (Author)
This work is licensed under a Creative Commons Attribution 4.0 International License.
