FPGA-QNN: Quantized Neural Network Hardware Acceleration on FPGAs

Recently, convolutional neural networks (CNNs) have received a massive amount of interest due to their ability to achieve high accuracy in various artificial intelligence tasks. With the development of complex CNN models, a significant drawback is their high computational burden and memory requireme...

Full description

Saved in:
Bibliographic Details
Main Authors: Mustafa Tasci, Ayhan Istanbullu, Vedat Tumen, Selahattin Kosunalp
Format: Article
Language:English
Published: MDPI AG 2025-01-01
Series:Applied Sciences
Subjects:
Online Access:https://www.mdpi.com/2076-3417/15/2/688
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1832589285515067392
author Mustafa Tasci
Ayhan Istanbullu
Vedat Tumen
Selahattin Kosunalp
author_facet Mustafa Tasci
Ayhan Istanbullu
Vedat Tumen
Selahattin Kosunalp
author_sort Mustafa Tasci
collection DOAJ
description Recently, convolutional neural networks (CNNs) have received a massive amount of interest due to their ability to achieve high accuracy in various artificial intelligence tasks. With the development of complex CNN models, a significant drawback is their high computational burden and memory requirements. The performance of a typical CNN model can be enhanced by the improvement of hardware accelerators. Practical implementations on field-programmable gate arrays (FPGA) have the potential to reduce resource utilization while maintaining low power consumption. Nevertheless, when implementing complex CNN models on FPGAs, these may may require further computational and memory capacities, exceeding the available capacity provided by many current FPGAs. An effective solution to this issue is to use quantized neural network (QNN) models to remove the burden of full-precision weights and activations. This article proposes an accelerator design framework for FPGAs, called FPGA-QNN, with a particular value in reducing high computational burden and memory requirements when implementing CNNs. To approach this goal, FPGA-QNN exploits the basics of quantized neural network (QNN) models by converting the high burden of full-precision weights and activations into integer operations. The FPGA-QNN framework comes up with 12 accelerators based on multi-layer perceptron (MLP) and LeNet CNN models, each of which is associated with a specific combination of quantization and folding. The outputs from the performance evaluations on Xilinx PYNQ Z1 development board proved the superiority of FPGA-QNN in terms of resource utilization and energy efficiency in comparison to several recent approaches. The proposed MLP model classified the FashionMNIST dataset at a speed of 953 kFPS with 1019 GOPs while consuming 2.05 W.
format Article
id doaj-art-b524110d1cbf4a349f667b017f797c80
institution Kabale University
issn 2076-3417
language English
publishDate 2025-01-01
publisher MDPI AG
record_format Article
series Applied Sciences
spelling doaj-art-b524110d1cbf4a349f667b017f797c802025-01-24T13:20:28ZengMDPI AGApplied Sciences2076-34172025-01-0115268810.3390/app15020688FPGA-QNN: Quantized Neural Network Hardware Acceleration on FPGAsMustafa Tasci0Ayhan Istanbullu1Vedat Tumen2Selahattin Kosunalp3Department of Computer Technologies, Gönen Vocational School, Bandırma Onyedi Eylül University, Bandırma 10200, TürkiyeDepartment of Computer Engineering, Faculty of Engineering, Balıkesir University, Balıkesir 10145, TürkiyeDepartment of Computer Engineering, Faculty of Engineering and Architecture, Bitlis Eren University, Bitlis 13100, TürkiyeDepartment of Computer Technologies, Gönen Vocational School, Bandırma Onyedi Eylül University, Bandırma 10200, TürkiyeRecently, convolutional neural networks (CNNs) have received a massive amount of interest due to their ability to achieve high accuracy in various artificial intelligence tasks. With the development of complex CNN models, a significant drawback is their high computational burden and memory requirements. The performance of a typical CNN model can be enhanced by the improvement of hardware accelerators. Practical implementations on field-programmable gate arrays (FPGA) have the potential to reduce resource utilization while maintaining low power consumption. Nevertheless, when implementing complex CNN models on FPGAs, these may may require further computational and memory capacities, exceeding the available capacity provided by many current FPGAs. An effective solution to this issue is to use quantized neural network (QNN) models to remove the burden of full-precision weights and activations. This article proposes an accelerator design framework for FPGAs, called FPGA-QNN, with a particular value in reducing high computational burden and memory requirements when implementing CNNs. To approach this goal, FPGA-QNN exploits the basics of quantized neural network (QNN) models by converting the high burden of full-precision weights and activations into integer operations. The FPGA-QNN framework comes up with 12 accelerators based on multi-layer perceptron (MLP) and LeNet CNN models, each of which is associated with a specific combination of quantization and folding. The outputs from the performance evaluations on Xilinx PYNQ Z1 development board proved the superiority of FPGA-QNN in terms of resource utilization and energy efficiency in comparison to several recent approaches. The proposed MLP model classified the FashionMNIST dataset at a speed of 953 kFPS with 1019 GOPs while consuming 2.05 W.https://www.mdpi.com/2076-3417/15/2/688acceleratorFPGAQNNdeep learningFINN
spellingShingle Mustafa Tasci
Ayhan Istanbullu
Vedat Tumen
Selahattin Kosunalp
FPGA-QNN: Quantized Neural Network Hardware Acceleration on FPGAs
Applied Sciences
accelerator
FPGA
QNN
deep learning
FINN
title FPGA-QNN: Quantized Neural Network Hardware Acceleration on FPGAs
title_full FPGA-QNN: Quantized Neural Network Hardware Acceleration on FPGAs
title_fullStr FPGA-QNN: Quantized Neural Network Hardware Acceleration on FPGAs
title_full_unstemmed FPGA-QNN: Quantized Neural Network Hardware Acceleration on FPGAs
title_short FPGA-QNN: Quantized Neural Network Hardware Acceleration on FPGAs
title_sort fpga qnn quantized neural network hardware acceleration on fpgas
topic accelerator
FPGA
QNN
deep learning
FINN
url https://www.mdpi.com/2076-3417/15/2/688
work_keys_str_mv AT mustafatasci fpgaqnnquantizedneuralnetworkhardwareaccelerationonfpgas
AT ayhanistanbullu fpgaqnnquantizedneuralnetworkhardwareaccelerationonfpgas
AT vedattumen fpgaqnnquantizedneuralnetworkhardwareaccelerationonfpgas
AT selahattinkosunalp fpgaqnnquantizedneuralnetworkhardwareaccelerationonfpgas