TY - JOUR
T1 - Nanophotonic Cavity Based Synapse for Scalable Photonic Neural Networks
AU - Jha, Aashu
AU - Huang, Chaoran
AU - Delima, Thomas Ferreira
AU - Peng, Hsuan Tung
AU - Shastri, Bhavin
AU - Prucnal, Paul R.
N1 - Publisher Copyright:
© 1995-2012 IEEE.
PY - 2022
Y1 - 2022
N2 - The bandwidth and energy demands of neural networks has spurred tremendous interest in developing novel neuromorphic hardware, including photonic integrated circuits. Although an optical waveguide can accommodate hundreds of channels with THz bandwidth, the channel count of photonic systems is always bottlenecked by the devices within. In WDM-based photonic neural networks, the synapses, i.e. network interconnections, are typically realized by microring resonators (MRRs), where the WDM channel count (N) is bounded by the free-spectral range of the MRRs. For typical Si MRRs, we estimate N ≤ 30 within the C-band. This not only restrains the aggregate throughput of the neural network but also makes applications with high input dimensions unfeasible. We experimentally demonstrate that photonic crystal nanobeam based synapses can be FSR-free within C-band, eliminating the bound on channel count. This increases data throughput as well as enables applications with high-dimensional inputs like natural language processing and high resolution image processing. In addition, the smaller physical footprint of photonic crystal nanobeam cavities offers higher tuning energy efficiency and a higher compute density than MRRs. Nanophotonic cavity based synapse thus offers a path towards realizing highly scalable photonic neural networks.
AB - The bandwidth and energy demands of neural networks has spurred tremendous interest in developing novel neuromorphic hardware, including photonic integrated circuits. Although an optical waveguide can accommodate hundreds of channels with THz bandwidth, the channel count of photonic systems is always bottlenecked by the devices within. In WDM-based photonic neural networks, the synapses, i.e. network interconnections, are typically realized by microring resonators (MRRs), where the WDM channel count (N) is bounded by the free-spectral range of the MRRs. For typical Si MRRs, we estimate N ≤ 30 within the C-band. This not only restrains the aggregate throughput of the neural network but also makes applications with high input dimensions unfeasible. We experimentally demonstrate that photonic crystal nanobeam based synapses can be FSR-free within C-band, eliminating the bound on channel count. This increases data throughput as well as enables applications with high-dimensional inputs like natural language processing and high resolution image processing. In addition, the smaller physical footprint of photonic crystal nanobeam cavities offers higher tuning energy efficiency and a higher compute density than MRRs. Nanophotonic cavity based synapse thus offers a path towards realizing highly scalable photonic neural networks.
KW - Photonic integrated circuits
KW - photonic neural networks
UR - http://www.scopus.com/inward/record.url?scp=85131720287&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85131720287&partnerID=8YFLogxK
U2 - 10.1109/JSTQE.2022.3179983
DO - 10.1109/JSTQE.2022.3179983
M3 - Article
AN - SCOPUS:85131720287
SN - 1077-260X
VL - 28
JO - IEEE Journal of Selected Topics in Quantum Electronics
JF - IEEE Journal of Selected Topics in Quantum Electronics
IS - 6
M1 - 6100908
ER -