Numerical simulation of an InP photonic integrated cross-connect for deep neural networks on chip

Bin Shi (Corresponding author), Nicola Calabretta, Ripalta Stabile

Onderzoeksoutput: Bijdrage aan tijdschriftTijdschriftartikelAcademicpeer review

7 Downloads (Pure)

Uittreksel

We propose a novel photonic accelerator architecture based on a broadcast-and-weight approach for a deep neural network through a photonic integrated cross-connect. The single neuron and the complete neural network operation are numerically simulated. The weight calibration and weighted addition are reproduced and demonstrated to behave as in the experimental measurements. A dynamic range higher than 25 dB is predicted, in line with the measurements. The weighted addition operation is also simulated and analyzed as a function of the optical crosstalk and the number of input colors involved. In particular, while an increase in optical crosstalk negatively influences the simulated error, a greater number of channels results in better performance. The iris flower classification problem is solved by implementing the weight matrix of a trained three-layer deep neural network. The performance of the corresponding photonic implementation is numerically investigated by tuning the optical crosstalk and waveguide loss, in order to anticipate energy consumption per operation. The analysis of the prediction error as a function of the optical crosstalk per layer suggests that the first layer is essential to the final accuracy. The ultimate accuracy shows a quasi-linear dependence between the prediction accuracy and the errors per layer for a normalized root mean square error lower than 0.09, suggesting that there is a maximum level of error permitted at the first layer for guaranteeing a final accuracy higher than 89%. However, it is still possible to find good local minima even for an error higher than 0.09, due to the stochastic nature of the network we are analyzing. Lower levels of path losses allow for half the power consumption at the matrix multiplication unit, for the same error level, offering opportunities for further improved performance. The good agreement between the simulations and the experiments offers a solid base for studying the scalability of this kind of network.
Originele taal-2Engels
Artikelnummer474
Aantal pagina's15
TijdschriftApplied Sciences
Volume10
Nummer van het tijdschrift2
DOI's
StatusGepubliceerd - 9 jan 2020

Vingerafdruk

Photonics
chips
photonics
Crosstalk
crosstalk
Computer simulation
simulation
Electric losses
root-mean-square errors
energy consumption
matrices
predictions
neurons
multiplication
Mean square error
Neurons
Particle accelerators
dynamic range
Deep neural networks
Network-on-chip

Citeer dit

@article{4c0d5699be5f4a7c863f3b4caf641a29,
title = "Numerical simulation of an InP photonic integrated cross-connect for deep neural networks on chip",
abstract = "We propose a novel photonic accelerator architecture based on a broadcast-and-weight approach for a deep neural network through a photonic integrated cross-connect. The single neuron and the complete neural network operation are numerically simulated. The weight calibration and weighted addition are reproduced and demonstrated to behave as in the experimental measurements. A dynamic range higher than 25 dB is predicted, in line with the measurements. The weighted addition operation is also simulated and analyzed as a function of the optical crosstalk and the number of input colors involved. In particular, while an increase in optical crosstalk negatively influences the simulated error, a greater number of channels results in better performance. The iris flower classification problem is solved by implementing the weight matrix of a trained three-layer deep neural network. The performance of the corresponding photonic implementation is numerically investigated by tuning the optical crosstalk and waveguide loss, in order to anticipate energy consumption per operation. The analysis of the prediction error as a function of the optical crosstalk per layer suggests that the first layer is essential to the final accuracy. The ultimate accuracy shows a quasi-linear dependence between the prediction accuracy and the errors per layer for a normalized root mean square error lower than 0.09, suggesting that there is a maximum level of error permitted at the first layer for guaranteeing a final accuracy higher than 89{\%}. However, it is still possible to find good local minima even for an error higher than 0.09, due to the stochastic nature of the network we are analyzing. Lower levels of path losses allow for half the power consumption at the matrix multiplication unit, for the same error level, offering opportunities for further improved performance. The good agreement between the simulations and the experiments offers a solid base for studying the scalability of this kind of network.",
author = "Bin Shi and Nicola Calabretta and Ripalta Stabile",
year = "2020",
month = "1",
day = "9",
doi = "10.3390/app10020474",
language = "English",
volume = "10",
journal = "Applied Sciences",
issn = "2076-3417",
publisher = "Multidisciplinary Digital Publishing Institute (MDPI)",
number = "2",

}

Numerical simulation of an InP photonic integrated cross-connect for deep neural networks on chip. / Shi, Bin (Corresponding author); Calabretta, Nicola; Stabile, Ripalta .

In: Applied Sciences, Vol. 10, Nr. 2, 474, 09.01.2020.

Onderzoeksoutput: Bijdrage aan tijdschriftTijdschriftartikelAcademicpeer review

TY - JOUR

T1 - Numerical simulation of an InP photonic integrated cross-connect for deep neural networks on chip

AU - Shi, Bin

AU - Calabretta, Nicola

AU - Stabile, Ripalta

PY - 2020/1/9

Y1 - 2020/1/9

N2 - We propose a novel photonic accelerator architecture based on a broadcast-and-weight approach for a deep neural network through a photonic integrated cross-connect. The single neuron and the complete neural network operation are numerically simulated. The weight calibration and weighted addition are reproduced and demonstrated to behave as in the experimental measurements. A dynamic range higher than 25 dB is predicted, in line with the measurements. The weighted addition operation is also simulated and analyzed as a function of the optical crosstalk and the number of input colors involved. In particular, while an increase in optical crosstalk negatively influences the simulated error, a greater number of channels results in better performance. The iris flower classification problem is solved by implementing the weight matrix of a trained three-layer deep neural network. The performance of the corresponding photonic implementation is numerically investigated by tuning the optical crosstalk and waveguide loss, in order to anticipate energy consumption per operation. The analysis of the prediction error as a function of the optical crosstalk per layer suggests that the first layer is essential to the final accuracy. The ultimate accuracy shows a quasi-linear dependence between the prediction accuracy and the errors per layer for a normalized root mean square error lower than 0.09, suggesting that there is a maximum level of error permitted at the first layer for guaranteeing a final accuracy higher than 89%. However, it is still possible to find good local minima even for an error higher than 0.09, due to the stochastic nature of the network we are analyzing. Lower levels of path losses allow for half the power consumption at the matrix multiplication unit, for the same error level, offering opportunities for further improved performance. The good agreement between the simulations and the experiments offers a solid base for studying the scalability of this kind of network.

AB - We propose a novel photonic accelerator architecture based on a broadcast-and-weight approach for a deep neural network through a photonic integrated cross-connect. The single neuron and the complete neural network operation are numerically simulated. The weight calibration and weighted addition are reproduced and demonstrated to behave as in the experimental measurements. A dynamic range higher than 25 dB is predicted, in line with the measurements. The weighted addition operation is also simulated and analyzed as a function of the optical crosstalk and the number of input colors involved. In particular, while an increase in optical crosstalk negatively influences the simulated error, a greater number of channels results in better performance. The iris flower classification problem is solved by implementing the weight matrix of a trained three-layer deep neural network. The performance of the corresponding photonic implementation is numerically investigated by tuning the optical crosstalk and waveguide loss, in order to anticipate energy consumption per operation. The analysis of the prediction error as a function of the optical crosstalk per layer suggests that the first layer is essential to the final accuracy. The ultimate accuracy shows a quasi-linear dependence between the prediction accuracy and the errors per layer for a normalized root mean square error lower than 0.09, suggesting that there is a maximum level of error permitted at the first layer for guaranteeing a final accuracy higher than 89%. However, it is still possible to find good local minima even for an error higher than 0.09, due to the stochastic nature of the network we are analyzing. Lower levels of path losses allow for half the power consumption at the matrix multiplication unit, for the same error level, offering opportunities for further improved performance. The good agreement between the simulations and the experiments offers a solid base for studying the scalability of this kind of network.

U2 - 10.3390/app10020474

DO - 10.3390/app10020474

M3 - Article

VL - 10

JO - Applied Sciences

JF - Applied Sciences

SN - 2076-3417

IS - 2

M1 - 474

ER -