An Efficient Frequency Encoding Scheme for Optical Convolution Accelerator
In today’s era where the demand for computational resources by large models is increasingly high, optical computing offers an alternative physical platform for computation. With its high parallelism and the maturation of integrated photonic technologies, optical computing is expected to further supp...
Saved in:
Main Authors: | , , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2024-12-01
|
Series: | Photonics |
Subjects: | |
Online Access: | https://www.mdpi.com/2304-6732/12/1/26 |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
_version_ | 1832587709492756480 |
---|---|
author | Gongyu Xia Jiacheng Liu Qilin Hong Pingyu Zhu Ping Xu Zhihong Zhu |
author_facet | Gongyu Xia Jiacheng Liu Qilin Hong Pingyu Zhu Ping Xu Zhihong Zhu |
author_sort | Gongyu Xia |
collection | DOAJ |
description | In today’s era where the demand for computational resources by large models is increasingly high, optical computing offers an alternative physical platform for computation. With its high parallelism and the maturation of integrated photonic technologies, optical computing is expected to further support the computational resources required by large models. For one-dimensional optical convolution accelerators, existing methods can fully utilize the working bandwidth of electro-optic modulators and the frequency resources of light sources. However, most convolution computations currently require the execution of two-dimensional matrix convolutions, and existing encoding schemes suffer a drop in terms of effective computations per second when performing two-dimensional matrix convolutions. In response to this, we propose a new encoding scheme that can fully utilize the computational resources of optical convolution accelerators. For convolution operations with a kernel size of <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><mrow><mi>M</mi><mo>×</mo><mi>M</mi></mrow></semantics></math></inline-formula>, compared to existing encoding schemes, it can achieve an <i>M</i>-fold increase in effective computation rate. This implies that under the constraints of essentially the same hardware physical performance, adopting our newly proposed encoding scheme can achieve a significant improvement in computational performance. We construct an optical convolution acceleration system to demonstrate the principles of the new encoding. |
format | Article |
id | doaj-art-5c7d547bb06948c09d0da7c8f8fa1252 |
institution | Kabale University |
issn | 2304-6732 |
language | English |
publishDate | 2024-12-01 |
publisher | MDPI AG |
record_format | Article |
series | Photonics |
spelling | doaj-art-5c7d547bb06948c09d0da7c8f8fa12522025-01-24T13:46:14ZengMDPI AGPhotonics2304-67322024-12-011212610.3390/photonics12010026An Efficient Frequency Encoding Scheme for Optical Convolution AcceleratorGongyu Xia0Jiacheng Liu1Qilin Hong2Pingyu Zhu3Ping Xu4Zhihong Zhu5College of Advanced Interdisciplinary Studies & Hunan Provincial Key Laboratory of Novel Nano Optoelectronic Information Materials and Devices, National University of Defense Technology, Changsha 410073, ChinaCollege of Advanced Interdisciplinary Studies & Hunan Provincial Key Laboratory of Novel Nano Optoelectronic Information Materials and Devices, National University of Defense Technology, Changsha 410073, ChinaCollege of Advanced Interdisciplinary Studies & Hunan Provincial Key Laboratory of Novel Nano Optoelectronic Information Materials and Devices, National University of Defense Technology, Changsha 410073, ChinaInstitute for Quantum Information and State Key Laboratory of High-Performance Computing, College of Computer Science and Technology, National University of Defense Technology, Changsha 410073, ChinaInstitute for Quantum Information and State Key Laboratory of High-Performance Computing, College of Computer Science and Technology, National University of Defense Technology, Changsha 410073, ChinaCollege of Advanced Interdisciplinary Studies & Hunan Provincial Key Laboratory of Novel Nano Optoelectronic Information Materials and Devices, National University of Defense Technology, Changsha 410073, ChinaIn today’s era where the demand for computational resources by large models is increasingly high, optical computing offers an alternative physical platform for computation. With its high parallelism and the maturation of integrated photonic technologies, optical computing is expected to further support the computational resources required by large models. For one-dimensional optical convolution accelerators, existing methods can fully utilize the working bandwidth of electro-optic modulators and the frequency resources of light sources. However, most convolution computations currently require the execution of two-dimensional matrix convolutions, and existing encoding schemes suffer a drop in terms of effective computations per second when performing two-dimensional matrix convolutions. In response to this, we propose a new encoding scheme that can fully utilize the computational resources of optical convolution accelerators. For convolution operations with a kernel size of <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><mrow><mi>M</mi><mo>×</mo><mi>M</mi></mrow></semantics></math></inline-formula>, compared to existing encoding schemes, it can achieve an <i>M</i>-fold increase in effective computation rate. This implies that under the constraints of essentially the same hardware physical performance, adopting our newly proposed encoding scheme can achieve a significant improvement in computational performance. We construct an optical convolution acceleration system to demonstrate the principles of the new encoding.https://www.mdpi.com/2304-6732/12/1/26convolution neural networkoptical computingmatrix convolution |
spellingShingle | Gongyu Xia Jiacheng Liu Qilin Hong Pingyu Zhu Ping Xu Zhihong Zhu An Efficient Frequency Encoding Scheme for Optical Convolution Accelerator Photonics convolution neural network optical computing matrix convolution |
title | An Efficient Frequency Encoding Scheme for Optical Convolution Accelerator |
title_full | An Efficient Frequency Encoding Scheme for Optical Convolution Accelerator |
title_fullStr | An Efficient Frequency Encoding Scheme for Optical Convolution Accelerator |
title_full_unstemmed | An Efficient Frequency Encoding Scheme for Optical Convolution Accelerator |
title_short | An Efficient Frequency Encoding Scheme for Optical Convolution Accelerator |
title_sort | efficient frequency encoding scheme for optical convolution accelerator |
topic | convolution neural network optical computing matrix convolution |
url | https://www.mdpi.com/2304-6732/12/1/26 |
work_keys_str_mv | AT gongyuxia anefficientfrequencyencodingschemeforopticalconvolutionaccelerator AT jiachengliu anefficientfrequencyencodingschemeforopticalconvolutionaccelerator AT qilinhong anefficientfrequencyencodingschemeforopticalconvolutionaccelerator AT pingyuzhu anefficientfrequencyencodingschemeforopticalconvolutionaccelerator AT pingxu anefficientfrequencyencodingschemeforopticalconvolutionaccelerator AT zhihongzhu anefficientfrequencyencodingschemeforopticalconvolutionaccelerator AT gongyuxia efficientfrequencyencodingschemeforopticalconvolutionaccelerator AT jiachengliu efficientfrequencyencodingschemeforopticalconvolutionaccelerator AT qilinhong efficientfrequencyencodingschemeforopticalconvolutionaccelerator AT pingyuzhu efficientfrequencyencodingschemeforopticalconvolutionaccelerator AT pingxu efficientfrequencyencodingschemeforopticalconvolutionaccelerator AT zhihongzhu efficientfrequencyencodingschemeforopticalconvolutionaccelerator |