Survey of Deep Learning Neural Networks Implementation on FPGAs

Deep learning has recently indicated that FPGAs (Field-Programmable Gate Arrays) play a significant role in accelerating DLNNs (Deep Learning Neural Networks). The initial specification of DLNN is usually done using a high-level language such as python, followed by a manual transformation to HDL (Hardware Description Language) for synthesis using a vendor tool. This transformation is tedious and needs HDL expertise, which limits the relevance of FPGAs. This paper presents an updated survey of the existing frameworks for mapping DLNNs onto FPGAs, comparing their characteristics, architectural choices, and achieved performance. Besides, we provide a comprehensive evaluation of different tools and their effectiveness for mapping DLNNs onto FPGAs. Finally, we present the future works.

Computer Science Engineering

View graph Source

Paper

Full text

Survey of Deep Learning Neural Networks Implementation on FPGAs

Semantic Scholar · Computer Science · 2020

Abstract

Deep learning has recently indicated that FPGAs (Field-Programmable Gate Arrays) play a significant role in accelerating DLNNs (Deep Learning Neural Networks). The initial specification of DLNN is usually done using a high-level language such as python, followed by a manual transformation to HDL (Hardware Description Language) for synthesis using a vendor tool. This transformation is tedious and needs HDL expertise, which limits the relevance of FPGAs. This paper presents an updated survey of the existing frameworks for mapping DLNNs onto FPGAs, comparing their characteristics, architectural choices, and achieved performance. Besides, we provide a comprehensive evaluation of different tools and their effectiveness for mapping DLNNs onto FPGAs. Finally, we present the future works.

Source of record · Semantic Scholar

Similar papers

Field-Programmable Gate Array Architecture for Deep Learning: Survey and Future DirectionsAndrew Boutros, Aman Arora, V. Betz2024 · Proceedings of the IEEE · 28 citations In Library

Accelerating Deep Neural Networks implementation: A surveyMeriam Dhouibi, Ahmed Karim Ben Salem, Afef Saidi et al.2021 · IET Computers & Digital Techniques · 26 citations In Library

Field-Programmable Gate Array Architecture for Deep Learning: Survey and Future DirectionsAndrew Boutros, Aman Arora, Vaughn Betz2025 · Proceedings of the IEEE In Library

A Survey and Taxonomy of FPGA-based Deep Learning AcceleratorsAhmed Ghazi Blaiech, K. Khalifa, C. Valderrama et al.2019 · Journal of systems architecture · 75 citations In Library

Implementation and Applications of Neural Networks Based on FPGAJiajun Li, Chuhuan Zhang, Xiangcheng Ma2024 In Library

An Automated Tool for Implementing Deep Neural Networks on FPGAMasoud Shahshahani, Mohammad Omar Sabri, Bahareh Khabbazan et al.2021 · International Conference on VLSI Design · 5 citations In Library

Deep Learning on FPGAs: Past, Present, and FutureGriffin Lacey, Graham W. Taylor, S. Areibi2016 · arXiv.org · 183 citations In Library

A Survey on LUT-based Deep Neural Networks Implemented in FPGAsZeyu Guo2025 · arXiv (Cornell University) · 6 citations In Library

A Survey of FPGA Based Deep Learning Accelerators: Challenges and OpportunitiesTeng Wang, Chao Wang, Xuehai Zhou et al.2018 · arXiv.org · 36 citations In Library

Raising the Abstraction Level of a Deep Learning Design on FPGAsDarío Baptista, L. Sousa, F. Morgado‐Dias2020 · IEEE Access · 6 citations In Library

Source of record

Semantic Scholar

44B

Welcome.

This panel sticks with you. Pick anything below and it opens right beside it, so you can dig through 60,000-plus records without ever losing your spot here.

Travel 44B

LibraryPapers, policy, standards, statute LabsEvery organization building AI ModelsIntelligence, price, and speed BenchmarksThe evaluation catalog IncidentsAI failures and harms SearchOne field across everything DashboardThe state of AI accountability in NY 44B RegistryThe Art. 44-B compliance portal