Journal of Applied Science, Engineering, Technology, and Education
Vol. 6 No. 1 (2024)

Design of Quantized Deep Neural Network Hardware Inference Accelerator Using Systolic Architecture

Dary Mochamad Rifqie (Department of Electronics Engineering, Universitas Negeri Makasssar, Makassar, 90223, Indonesia)
Yasser Abd. Djawad (Department of Electronics Engineering, Universitas Negeri Makasssar, Makassar, 90223, Indonesia)
Faizal Arya Samman (Department of Electrical Engineering, Universitas Hasanuddin, Makassar, 90245, Indonesia)
Ansari Saleh Ahmar (Department of Statistics, Universitas Negeri Makassar, Makassar, 90223, Indonesia)
M. Miftach Fakhri (Department of Informatics and Computer Engineering Education, Universitas Negeri Makasssar, Makassar, 90223, Indonesia)

Article Info

Publish Date
28 Jun 2024

Abstract

This paper presents a hardware inference accelerator architecture of quantized deep neural networks (DNN). The proposed accelerator implements all computation in a quantize version of DNN including linear transformations like matrix multiplications, nonlinear activation functions such as ReLU, quantization and dequantization operation. The hardware accelerator of quantized DNN consists of matrix multiplication core which is implemented in systolic array architecture, and the QDR core for computing the operation of quantization, dequantization, and ReLU. This proposed hardware architecture is implemented in Verilog Hardware Description Language (HDL) code using modelsim. To validate, we simulated the quantized DNN using Python programming language and compared the results with our proposed hardware accelerator. The result of this comparison shows a very slight difference, confirming the validity of our quantized DNN hardware accelerator.

Copyrights © 2024




Citation Download
RIS
EndNote, Reference Manager, ProCite
BibTex
Latex, Jabref
Original Source
Download Original
Google Scholar
Check in Google Scholar
Journal Info
Journal of Applied Science, Engineering, Technology, and Education
Website

Abbrev

asci

Publisher

PT. Mattawang Mediatama Solution

Subject

Civil Engineering, Building, Construction & Architecture Computer Science & IT Electrical & Electronics Engineering Industrial & Manufacturing Engineering Other

Description

Journal of Applied Science, Engineering, Technology, and Education (ASCI) is an international wide scope, peer-reviewed open access journal for the publication of original papers concerned with diverse aspects of science application, technology and ...