Large-scale integrated photonic accelerators for ultralow-latency and universal AI computing

Xiangyan Meng; Junshen Li; Kangwei Fei; Yu Wang; Wei Li; Nuannuan Shi; Ming Li

doi:10.1088/1674-4926/26020057

J. Semicond. > Just Accepted

NEWS AND VIEWS

Large-scale integrated photonic accelerators for ultralow-latency and universal AI computing

Xiangyan Meng^{1, 2, 3}, Junshen Li^{1, 2, 3}, Kangwei Fei^{1, 2, 3}, Yu Wang^{1, 2, 3}, Wei Li^{1, 2, 3}, Nuannuan Shi^{1, 2, 3,} and Ming Li^{1, 2, 3,}

+ Author Affiliations

1. State Key Laboratory of Optoelectronic Materials and Devices, Institute of Semiconductors, Chinese Academy of Sciences, Beijing 100083, China
2. College of Materials Science and Opto-Electronic Technology, University of Chinese Academy of Sciences, Beijing 100049, China
3. School of Electronic, Electrical and Communication Engineering, University of Chinese Academy of Sciences, Beijing 100049, China

Author Bio:

Xiangyan Meng received the B.Eng. degree from the College of Science, China University of Petroleum, Qingdao, China, in 2019 and the Ph.D. degree from Institute of Semiconductors, CAS in 2024. He is currently a postdoctoral fellow at the Institute of Semiconductors, CAS, Beijing, China. His research interests include intelligent optical computing and photonics integration.

Nuannuan Shi received the Ph.D. degree in optical engineering from the Dalian University of Technology, Dalian, China, in 2015. From 2015 to 2017, she was with the Institute of Semiconductors, CAS, Beijing, as a Postdoctoral Research Fellow. She is currently an Professor with the Institute of Semiconductors, CAS. She has published more than 20 papers in Nature Communication, IEEE Journal of Lightwave Technology, Optics Express, etc. Her research interests include integrated microwave photonics and its applications, optical computing, and signal processing.

Ming Li (Fellow, Optica) received the Ph.D. degree in electrical and electronics engineering from the University of Shizuoka, Hamamatsu, Japan, in 2009. In 2009, he was with the Microwave Photonics Research Laboratory, School of Electrical Engineering and Computer Science, University of Ottawa, Ottawa, ON, Canada, as a Postdoctoral Research Fellow. In 2011, he was in the Ultrafast Optical Processing Group under the supervision of INRS-EMT, Montreal, QC, Canada, as a Postdoctoral Research Fellow. In 2013, he was with the Institute of Semiconductors, CAS, as a Full Professor under the support of Thousand Youth Talents Program. He has published more than 210 high-impact journal papers in Nature Photonics, Nature Communication, Light: Science & Applications, Physical Review Letters, etc. His research interests include integrated microwave photonics and its applications, ultrafast optical signal processing, and high-speed real-time optical measurement and sensing.

Corresponding author: Nuannuan Shi, nnshi@semi.ac.cn; Ming Li, ml@semi.ac.cn

DOI: 10.1088/1674-4926/26020057CSTR: 32376.14.1674-4926.26020057

References

[1]	Shastri B J, Tait A N, Ferreira de Lima T, et al. Photonics for artificial intelligence and neuromorphic computing. Nat Photonics, 2021, 15(2): 102 doi: 10.1038/s41566-020-00754-y
[2]	Shen Y C, Harris N C, Skirlo S, et al. Deep learning with coherent nanophotonic circuits. Nat Photonics, 2017, 11(7): 441 doi: 10.1038/nphoton.2017.93
[3]	Feldmann J, Youngblood N, Karpov M, et al. Parallel convolutional processing using an integrated photonic tensor core. Nature, 2021, 589(7840): 52 doi: 10.1038/s41586-020-03070-1
[4]	Xu X Y, Tan M X, Corcoran B, et al. 11 TOPS photonic convolutional accelerator for optical neural networks. Nature, 2021, 589(7840): 44 doi: 10.1038/s41586-020-03063-0
[5]	Dong B W, Brückerhoff-Plückelmann F, Meyer L, et al. Partial coherence enhances parallelized photonic computing. Nature, 2024, 632(8023): 55 doi: 10.1038/s41586-024-07590-y
[6]	Chen Z J, Sludds A, Davis R, et al. Deep learning with coherent VCSEL neural networks. Nat Photonics, 2023, 17(8): 723 doi: 10.1038/s41566-023-01233-w
[7]	Xu Z H, Zhou T K, Ma M Z, et al. Large-scale photonic chiplet Taichi empowers 160-TOPS/W artificial general intelligence. Science, 2024, 384(6692): 202 doi: 10.1126/science.adl1203
[8]	LeCun Y, Bengio Y, Hinton G. Deep learning. Nature, 2015, 521(7553): 436 doi: 10.1038/nature14539
[9]	Roques-Carmes C, Shen Y C, Zanoci C, et al. Heuristic recurrent algorithms for photonic Ising machines. Nat Commun, 2020, 11: 249 doi: 10.1038/s41467-019-14096-z
[10]	Pintus P, Dumont M, Shah V, et al. Integrated non-reciprocal magneto-optics with ultra-high endurance for photonic in-memory computing. Nat Photonics, 2025, 19(1): 54 doi: 10.1038/s41566-024-01549-1
[11]	Becker S, Englund D, Stiller B. An optoacoustic field-programmable perceptron for recurrent neural networks. Nat Commun, 2024, 15: 3020 doi: 10.1038/s41467-024-47053-6
[12]	Hua S Y, Divita E, Yu S S, et al. An integrated large-scale photonic accelerator with ultralow latency. Nature, 2025, 640(8058): 361 doi: 10.1038/s41586-025-08786-6
[13]	Ahmed S R, Baghdadi R, Bernadskiy M, et al. Universal photonic artificial intelligence acceleration. Nature, 2025, 640(8058): 368 doi: 10.1038/s41586-025-08854-x

Fig. 1. (Color online) Chip architecture, board-level photograph, and chip packaging schematic of PACE (a) and the universal photonic AI processor (b).

DownLoad: Full-Size Img PowerPoint

[1]	Shastri B J, Tait A N, Ferreira de Lima T, et al. Photonics for artificial intelligence and neuromorphic computing. Nat Photonics, 2021, 15(2): 102 doi: 10.1038/s41566-020-00754-y
[2]	Shen Y C, Harris N C, Skirlo S, et al. Deep learning with coherent nanophotonic circuits. Nat Photonics, 2017, 11(7): 441 doi: 10.1038/nphoton.2017.93
[3]	Feldmann J, Youngblood N, Karpov M, et al. Parallel convolutional processing using an integrated photonic tensor core. Nature, 2021, 589(7840): 52 doi: 10.1038/s41586-020-03070-1
[4]	Xu X Y, Tan M X, Corcoran B, et al. 11 TOPS photonic convolutional accelerator for optical neural networks. Nature, 2021, 589(7840): 44 doi: 10.1038/s41586-020-03063-0
[5]	Dong B W, Brückerhoff-Plückelmann F, Meyer L, et al. Partial coherence enhances parallelized photonic computing. Nature, 2024, 632(8023): 55 doi: 10.1038/s41586-024-07590-y
[6]	Chen Z J, Sludds A, Davis R, et al. Deep learning with coherent VCSEL neural networks. Nat Photonics, 2023, 17(8): 723 doi: 10.1038/s41566-023-01233-w
[7]	Xu Z H, Zhou T K, Ma M Z, et al. Large-scale photonic chiplet Taichi empowers 160-TOPS/W artificial general intelligence. Science, 2024, 384(6692): 202 doi: 10.1126/science.adl1203
[8]	LeCun Y, Bengio Y, Hinton G. Deep learning. Nature, 2015, 521(7553): 436 doi: 10.1038/nature14539
[9]	Roques-Carmes C, Shen Y C, Zanoci C, et al. Heuristic recurrent algorithms for photonic Ising machines. Nat Commun, 2020, 11: 249 doi: 10.1038/s41467-019-14096-z
[10]	Pintus P, Dumont M, Shah V, et al. Integrated non-reciprocal magneto-optics with ultra-high endurance for photonic in-memory computing. Nat Photonics, 2025, 19(1): 54 doi: 10.1038/s41566-024-01549-1
[11]	Becker S, Englund D, Stiller B. An optoacoustic field-programmable perceptron for recurrent neural networks. Nat Commun, 2024, 15: 3020 doi: 10.1038/s41467-024-47053-6
[12]	Hua S Y, Divita E, Yu S S, et al. An integrated large-scale photonic accelerator with ultralow latency. Nature, 2025, 640(8058): 361 doi: 10.1038/s41586-025-08786-6
[13]	Ahmed S R, Baghdadi R, Bernadskiy M, et al. Universal photonic artificial intelligence acceleration. Nature, 2025, 640(8058): 368 doi: 10.1038/s41586-025-08854-x

Search

GET CITATION

shu

Export: BibTex EndNote

Article Metrics

Article views: 16 Times PDF downloads: 4 Times Cited by: 0 Times

History

Received: 14 February 2026 Revised: Online: Accepted Manuscript: 30 March 2026

Article Navigation > Journal of Semiconductors > 2026 > Accepted Manuscript

Xiangyan Meng, Junshen Li, Kangwei Fei, Yu Wang, Wei Li, Nuannuan Shi, Ming Li. Large-scale integrated photonic accelerators for ultralow-latency and universal AI computing[J]. Journal of Semiconductors, 2026, In Press. doi: 10.1088/1674-4926/26020057 ****X Y Meng, J S Li, K W Fei, Y Wang, W Li, N N Shi, and M Li, Large-scale integrated photonic accelerators for ultralow-latency and universal AI computing[J]. J. Semicond., 2026, accepted doi: 10.1088/1674-4926/26020057

Citation:

X Y Meng, J S Li, K W Fei, Y Wang, W Li, N N Shi, and M Li, Large-scale integrated photonic accelerators for ultralow-latency and universal AI computing[J]. J. Semicond., 2026, accepted doi: 10.1088/1674-4926/26020057

Citation:

PDF( 2491 KB)

Large-scale integrated photonic accelerators for ultralow-latency and universal AI computing

DOI: 10.1088/1674-4926/26020057

CSTR: 32376.14.1674-4926.26020057

Xiangyan Meng^{1, 2, 3},
Junshen Li^{1, 2, 3},
Kangwei Fei^{1, 2, 3},
Yu Wang^{1, 2, 3},
Wei Li^{1, 2, 3},
Nuannuan Shi^{1, 2, 3,},
Ming Li^{1, 2, 3,}

1.
State Key Laboratory of Optoelectronic Materials and Devices, Institute of Semiconductors, Chinese Academy of Sciences, Beijing 100083, China
2.
College of Materials Science and Opto-Electronic Technology, University of Chinese Academy of Sciences, Beijing 100049, China
3.
School of Electronic, Electrical and Communication Engineering, University of Chinese Academy of Sciences, Beijing 100049, China

More Information

Xiangyan Meng received the B.Eng. degree from the College of Science, China University of Petroleum, Qingdao, China, in 2019 and the Ph.D. degree from Institute of Semiconductors, CAS in 2024. He is currently a postdoctoral fellow at the Institute of Semiconductors, CAS, Beijing, China. His research interests include intelligent optical computing and photonics integration
Nuannuan Shi received the Ph.D. degree in optical engineering from the Dalian University of Technology, Dalian, China, in 2015. From 2015 to 2017, she was with the Institute of Semiconductors, CAS, Beijing, as a Postdoctoral Research Fellow. She is currently an Professor with the Institute of Semiconductors, CAS. She has published more than 20 papers in Nature Communication, IEEE Journal of Lightwave Technology, Optics Express, etc. Her research interests include integrated microwave photonics and its applications, optical computing, and signal processing
Ming Li (Fellow, Optica) received the Ph.D. degree in electrical and electronics engineering from the University of Shizuoka, Hamamatsu, Japan, in 2009. In 2009, he was with the Microwave Photonics Research Laboratory, School of Electrical Engineering and Computer Science, University of Ottawa, Ottawa, ON, Canada, as a Postdoctoral Research Fellow. In 2011, he was in the Ultrafast Optical Processing Group under the supervision of INRS-EMT, Montreal, QC, Canada, as a Postdoctoral Research Fellow. In 2013, he was with the Institute of Semiconductors, CAS, as a Full Professor under the support of Thousand Youth Talents Program. He has published more than 210 high-impact journal papers in Nature Photonics, Nature Communication, Light: Science & Applications, Physical Review Letters, etc. His research interests include integrated microwave photonics and its applications, ultrafast optical signal processing, and high-speed real-time optical measurement and sensing
Corresponding author: nnshi@semi.ac.cn; ml@semi.ac.cn
Received Date: 2026-02-14
Available Online: 2026-03-30

Abstract

FullText(HTML)

References(13)