High-integrated photonic tensor core utilizing high-dimensional lightwave and microwave multidomain multiplexing
Abstract The burgeoning volume of parameters in artificial neural network models has posed substantial challenges to conventional tensor computing hardware. Benefiting from the available optical multidimensional information entropy, optical intelligent computing is used as an alternative solution to...
Saved in:
| Main Authors: | , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Publishing Group
2025-01-01
|
| Series: | Light: Science & Applications |
| Online Access: | https://doi.org/10.1038/s41377-024-01706-9 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1841559080414478336 |
|---|---|
| author | Xiangyan Meng Nuannuan Shi Guojie Zhang Junshen Li Ye Jin Shiyou Sun Yichen Shen Wei Li Ninghua Zhu Ming Li |
| author_facet | Xiangyan Meng Nuannuan Shi Guojie Zhang Junshen Li Ye Jin Shiyou Sun Yichen Shen Wei Li Ninghua Zhu Ming Li |
| author_sort | Xiangyan Meng |
| collection | DOAJ |
| description | Abstract The burgeoning volume of parameters in artificial neural network models has posed substantial challenges to conventional tensor computing hardware. Benefiting from the available optical multidimensional information entropy, optical intelligent computing is used as an alternative solution to address the emerging challenges of electrical computing. These limitations, in terms of device size and photonic integration scale, have hindered the performance of optical chips. Herein, an ultrahigh computing density optical tensor processing unit (OTPU), which is grounded in an individual microring resonator (MRR), is introduced to respond to these challenges. Through the independent tuning of multiwavelength lasers, the operational capabilities of an MRR are orchestrated, culminating in the formation of an optical tensor core. This design facilitates the execution of tensor convolution operations via the lightwave and microwave multidomain hybrid multiplexing in terms of the time, wavelength, and frequency of microwaves. The experimental results for the MRR-based OTPU show an extraordinary computing density of 34.04 TOPS/mm2. Additionally, the achieved accuracy rate in recognizing MNIST handwritten digits was 96.41%. These outcomes signify a significant advancement toward the realization of high-performance optical tensor processing chips. |
| format | Article |
| id | doaj-art-8ef3e5a0476941e69b41435bfaa66ece |
| institution | Kabale University |
| issn | 2047-7538 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | Nature Publishing Group |
| record_format | Article |
| series | Light: Science & Applications |
| spelling | doaj-art-8ef3e5a0476941e69b41435bfaa66ece2025-01-05T12:46:59ZengNature Publishing GroupLight: Science & Applications2047-75382025-01-0114111110.1038/s41377-024-01706-9High-integrated photonic tensor core utilizing high-dimensional lightwave and microwave multidomain multiplexingXiangyan Meng0Nuannuan Shi1Guojie Zhang2Junshen Li3Ye Jin4Shiyou Sun5Yichen Shen6Wei Li7Ninghua Zhu8Ming Li9Key Laboratory of Optoelectronic Materials and Devices, Institute of Semiconductors, Chinese Academy of SciencesKey Laboratory of Optoelectronic Materials and Devices, Institute of Semiconductors, Chinese Academy of SciencesChina Academy of Space Technology (Xi’an)Key Laboratory of Optoelectronic Materials and Devices, Institute of Semiconductors, Chinese Academy of SciencesKey Laboratory of Optoelectronic Materials and Devices, Institute of Semiconductors, Chinese Academy of SciencesWeChat Pay Lab 33, Shenzhen Tencent Computer System Co. Ltd.Lightelligence GroupKey Laboratory of Optoelectronic Materials and Devices, Institute of Semiconductors, Chinese Academy of SciencesKey Laboratory of Optoelectronic Materials and Devices, Institute of Semiconductors, Chinese Academy of SciencesKey Laboratory of Optoelectronic Materials and Devices, Institute of Semiconductors, Chinese Academy of SciencesAbstract The burgeoning volume of parameters in artificial neural network models has posed substantial challenges to conventional tensor computing hardware. Benefiting from the available optical multidimensional information entropy, optical intelligent computing is used as an alternative solution to address the emerging challenges of electrical computing. These limitations, in terms of device size and photonic integration scale, have hindered the performance of optical chips. Herein, an ultrahigh computing density optical tensor processing unit (OTPU), which is grounded in an individual microring resonator (MRR), is introduced to respond to these challenges. Through the independent tuning of multiwavelength lasers, the operational capabilities of an MRR are orchestrated, culminating in the formation of an optical tensor core. This design facilitates the execution of tensor convolution operations via the lightwave and microwave multidomain hybrid multiplexing in terms of the time, wavelength, and frequency of microwaves. The experimental results for the MRR-based OTPU show an extraordinary computing density of 34.04 TOPS/mm2. Additionally, the achieved accuracy rate in recognizing MNIST handwritten digits was 96.41%. These outcomes signify a significant advancement toward the realization of high-performance optical tensor processing chips.https://doi.org/10.1038/s41377-024-01706-9 |
| spellingShingle | Xiangyan Meng Nuannuan Shi Guojie Zhang Junshen Li Ye Jin Shiyou Sun Yichen Shen Wei Li Ninghua Zhu Ming Li High-integrated photonic tensor core utilizing high-dimensional lightwave and microwave multidomain multiplexing Light: Science & Applications |
| title | High-integrated photonic tensor core utilizing high-dimensional lightwave and microwave multidomain multiplexing |
| title_full | High-integrated photonic tensor core utilizing high-dimensional lightwave and microwave multidomain multiplexing |
| title_fullStr | High-integrated photonic tensor core utilizing high-dimensional lightwave and microwave multidomain multiplexing |
| title_full_unstemmed | High-integrated photonic tensor core utilizing high-dimensional lightwave and microwave multidomain multiplexing |
| title_short | High-integrated photonic tensor core utilizing high-dimensional lightwave and microwave multidomain multiplexing |
| title_sort | high integrated photonic tensor core utilizing high dimensional lightwave and microwave multidomain multiplexing |
| url | https://doi.org/10.1038/s41377-024-01706-9 |
| work_keys_str_mv | AT xiangyanmeng highintegratedphotonictensorcoreutilizinghighdimensionallightwaveandmicrowavemultidomainmultiplexing AT nuannuanshi highintegratedphotonictensorcoreutilizinghighdimensionallightwaveandmicrowavemultidomainmultiplexing AT guojiezhang highintegratedphotonictensorcoreutilizinghighdimensionallightwaveandmicrowavemultidomainmultiplexing AT junshenli highintegratedphotonictensorcoreutilizinghighdimensionallightwaveandmicrowavemultidomainmultiplexing AT yejin highintegratedphotonictensorcoreutilizinghighdimensionallightwaveandmicrowavemultidomainmultiplexing AT shiyousun highintegratedphotonictensorcoreutilizinghighdimensionallightwaveandmicrowavemultidomainmultiplexing AT yichenshen highintegratedphotonictensorcoreutilizinghighdimensionallightwaveandmicrowavemultidomainmultiplexing AT weili highintegratedphotonictensorcoreutilizinghighdimensionallightwaveandmicrowavemultidomainmultiplexing AT ninghuazhu highintegratedphotonictensorcoreutilizinghighdimensionallightwaveandmicrowavemultidomainmultiplexing AT mingli highintegratedphotonictensorcoreutilizinghighdimensionallightwaveandmicrowavemultidomainmultiplexing |