Coherent VCSEL Network Computing

Chen, Zaijun; Sludds, Alexander; Davis, Ronald; Christen, Ian; Ateshian, Lamia; Bernstein, Liane; Heuser, Tobias; Heermeier, Niels; Lott, James A.; Reitzenstein, Stephan; Hamerly, Ryan; Englund, Dirk

doi:10.23919/oecc/psc53152.2022.9849860

Cited by 6 publications

(2 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The neural network (NN) and other neuromorphic algorithms could be performed on these platforms in the analog domain through memristor (3,4), twodimensional (2D) materials (5)(6)(7) and other programmable components (8,9) with customized configurations and network architectures. Among them, photonic computing circuits open an avenue for direct optical signal processing with ultralow latency and unprecedented energy efficiency (10)(11)(12) and are considered as preferable processing units for large-scale neuromorphic applications in the future (13). Several network architectures have been proposed, such as diffractive deep NN (14), optical NN (ONN) (10), 2D photonic diffraction network (15)(16)(17)(18), and their derivative configurations for more complicated network structure realizations (19)(20)(21), with some practical applications (22,23).…”

Section: Introductionmentioning

confidence: 99%

Direct electromagnetic information processing with planar diffractive neural network

Gu,

Ma,

Gao

et al. 2024

Sci. Adv.

View full text Add to dashboard Cite

Diffractive neural network in electromagnetic wave–driven system has attracted great attention due to its ultrahigh parallel computing capability and energy efficiency. However, recent neural networks based on the diffractive framework still face the bottlenecks of misalignment and relatively large size limiting their further applications. Here, we propose a planar diffractive neural network (pla-NN) with a highly integrated and conformal architecture to achieve direct signal processing in the microwave frequency. On the basis of printed circuit fabrication process, the misalignment could be effectively circumvented while enabling flexible extension for multiple conformal and stacking designs. We first conduct validation on the fashion-MNIST dataset and experimentally build up a system using the proposed network architecture for direct recognition of different geometry structures in the electromagnetic space. We envision that the presented architecture, once combined with the advanced dynamic maneuvering techniques and flexible topology, would exhibit unlimited potentials in the areas of high-performance computing, wireless sensing, and flexible wearable electronics.

show abstract

Section: Introductionmentioning

confidence: 99%

Direct electromagnetic information processing with planar diffractive neural network

Gu,

Ma,

Gao

et al. 2024

Sci. Adv.

View full text Add to dashboard Cite

show abstract

“…In the field of deep neural networks, coherent VCSELs can be used to build optical neuronal networks. 6 Numerical simulation plays an important role in the design and optimization of these semiconductor lasers. 2,7 The device dimension with a diameter of 30 µm is large compared to the emission wavelength of λ = 980 nm, which implies a relatively large numerical effort for 3D optical simulations.…”

Section: Introductionmentioning

confidence: 99%

Applying a Riesz-projection-based contour integral eigenvalue solver to compute resonance modes of a VCSEL

Kuen

Betz

Binkowski

et al. 2023

Integrated Optics: Design, Devices, Systems and Applications VII

View full text Add to dashboard Cite

We investigate the calculation of resonance modes of a VCSEL with a Riesz projection eigenvalue solver. The eigenvalue solver is based on the principle of contour integration where for the solution of scattering problems physical right sides are used. Here, it is investigated how numerical parameters impact the performance of the method, where we focus on the computation of the fundamental VCSEL mode.

show abstract

Coherent VCSEL homodyne neural networks

Chen

Sludds

Davis

et al. 2023

AI and Optical Data Sciences IV

Self Cite

View full text Add to dashboard Cite

We demonstrated a large-scale space-time-multiplexed homodyne optical neural network (ONN) using arrays of highspeed (GHz) vertical-cavity surface-emitting lasers (VCSELs). Injection locking enables precise phase control over tens of VCSEL devices simultaneously, facilitating photoelectric-multiplication-based matrix operations and all-optical nonlinearity, operating at the quantum-noise limit. Our VCSEL transmitters exhibit ultra-high electro-optic conversion efficiency (V=4 mV), allowing neural encoding at 5 attojoule/symbol. Three-dimensional neural connectivity allows parallel computing. The full-system energy efficiency reaches 7 fJ/operation, which is >100-fold better than the state-ofthe-art digital microprocessors and other ONN demonstrations. Digit classification is achieved with an accuracy of 98% of the group truth.

show abstract

Coherent VCSEL Network Computing

Cited by 6 publications

References 12 publications

Direct electromagnetic information processing with planar diffractive neural network

Direct electromagnetic information processing with planar diffractive neural network

Applying a Riesz-projection-based contour integral eigenvalue solver to compute resonance modes of a VCSEL

Coherent VCSEL homodyne neural networks

Contact Info

Product

Resources

About