Library

Extending a Soft-Core RISC-V Processor to Accelerate CNN Inference

Convolutional Neural Networks (CNNs) are the gold-standard for computer vision. Using CNN on embedded hardware that has limited computational capability is an area of active investigation and optimization. In this paper, we investigate the potential of extending the RISC-V Instruction Set Architecture for accelerating the inference of a CNN using in-pipeline hardware blocks and custom instructions. Our preliminary designs have a small footprint and minimal impact on maximum core frequency. The new designed instructions were used to extend an existing soft-core processor. This processor was synthesized to an FPGA for cycle-accurate testing and performance evaluation.

Computer ScienceEngineering
View graphSource

Paper

Full text

PDF

Extending a Soft-Core RISC-V Processor to Accelerate CNN Inference

Semantic Scholar · Computer Science · 2019

Abstract

Convolutional Neural Networks (CNNs) are the gold-standard for computer vision. Using CNN on embedded hardware that has limited computational capability is an area of active investigation and optimization. In this paper, we investigate the potential of extending the RISC-V Instruction Set Architecture for accelerating the inference of a CNN using in-pipeline hardware blocks and custom instructions. Our preliminary designs have a small footprint and minimal impact on maximum core frequency. The new designed instructions were used to extend an existing soft-core processor. This processor was synthesized to an FPGA for cycle-accurate testing and performance evaluation.

Source of record · Semantic Scholar

Similar papers

CNN Specific ISA Extensions Based on RISC-V ProcessorsXiang Yu, Zhijie Yang, Linghui Peng et al.2022 In Library
A Reconfigurable Convolutional Neural Network-Accelerated Coprocessor Based on RISC-V Instruction SetNing Wu, Tao Jiang, Lei Zhang et al.2020 · Electronics · 35 citations In Library
Design of low-power acceleration processor for convolutional neural networks based on RISC-VYunfei Zhu, Xiao Zhang, Rongcai Zhao et al.2023 · 4th International Conference on Information Science, Electrical and Automation Engineering · 0 citations In Library
A CNN Accelerator with Embedded Risc-V ControllersLi Zhang, Xian Zhou, Chuliang Guo2021 In Library
Hybrid RISC-V Coprocessor Architecture for Convolutional Neural-Network InferencingAustin M. Richie, David A. Richie2026 In Library
Performance Evaluation of CNN using RISC-V Vector Extension for Embedded SystemManato Okawara, Kanemitsu Ootsu, Takuma Kitamoto et al.2025 In Library
Extensible Embedded Processor for Convolutional Neural NetworksJoshua Misko, Shrikant S. Jadhav, Youngsoo Kim2021 · Scientific Programming · 3 citations In Library
A Low-Power RISC-V Based Reconfigurable CNN Accelerator for On-Chip Inference in Wireless Communication Edge DevicesHelin Kong, Z. Wei, Tingran Chen et al.2025 In Library
Implementation of CNN Heterogeneous Scheme Based on Domestic FPGA with RISC-V Soft Core CPUHai‐Long Wu, Jindong Li, Xiang Chen2022 · 2022 IEEE International Conference on Integrated Circuits, Technologies and Applications (ICTA) In Library
Accelerating Machine Learning using RISC-V Vector Extension in a Manycore PlatformWillian Analdo Nunes, Antônio Vinicius Corrêa Dos Santos, César Marcon et al.2025 · IEEE/IFIP International Conference on Very Large Scale Integration of System-on-Chip · 0 citations In Library
© 2026 NYSGPT2525 LLC