CHAMP: Coherent Hardware-Aware Magnitude Pruning of Integrated Photonic Neural Networks

We propose a novel hardware-aware magnitude pruning technique for coherent photonic neural networks. The proposed technique can prune 99.45% of network parameters and reduce the static power consumption by 98.23% with a negligible accuracy loss.

cs.ET cs.LG Neural Networks and Reservoir Computing Optical Network Technologies Photonic and Optical Devices Engineering Physics Computer Science

View graph Source

Paper

Similar papers

CHAMP: Coherent Hardware-Aware Magnitude Pruning of Integrated Photonic Neural NetworksSanmitra Banerjee, Mahdi Nikdast, Sudeep Pasricha et al.2022 In Library

Pruning Coherent Integrated Photonic Neural NetworksSanmitra Banerjee, Mahdi Nikdast, Sudeep Pasricha et al.2023 · IEEE Journal of Selected Topics in Quantum Electronics In Library

Pruning Coherent Integrated Photonic Neural Networks Using the Lottery Ticket HypothesisSanmitra Banerjee, Mahdi Nikdast, Sudeep Pasricha et al.2022 In Library

Control-free and efficient integrated photonic neural networks via hardware-aware training and pruningTengji Xu, Weipeng Zhang, Jiawei Zhang et al.2024 · arXiv (Cornell University) · 31 citations In Library

Pruning Coherent Integrated Photonic Neural Networks Using the Lottery Ticket HypothesisSanmitra Banerjee, Mahdi Nikdast, Sudeep Pasricha et al.2021 · arXiv (Cornell University) · 11 citations In Library

Control-free and efficient integrated photonic neural networks via hardware-aware training and pruningTengji Xu, Weipeng Zhang, Jiawei Zhang et al.2024 · Optica In Library

Online training and pruning of multi-wavelength photonic neural networksJiawei Zhang, Weipeng Zhang, Tengji Xu et al.2024 · Nanophotonics · 2 citations In Library

Optimizing Coherent Integrated Photonic Neural Networks under Random UncertaintiesSanmitra Banerjee, Mahdi Nikdast, Krishnendu Chakrabarty2021 In Library

Online training and pruning of photonic neural networksWeipeng Zhang, Tengji Xu, Jiawei Zhang et al.2023 In Library

Power-aware pruning for ultrafast, energy-efficient, and accurate optical neural network designNaoki Hattori, Yutaka Masuda, Tohru Ishihara et al.2022 · Design Automation Conference · 0 citations In Library

Patents near this paper

SYSTEMS AND METHODS FOR NEURAL NETWORK TRAINING AND DEPLOYMENT FOR HARDWARE ACCELERATORSUS 11,544,540 · HEWLETT PACKARD ENTERPRISE DEVELOPMENT LP · 2023 Large-Scale Artificial Neural-Network Accelerators Based on Coherent Detection and Optical Data Fan-OutUS 11,604,978 · MASSACHUSETTS INSTITUTE OF TECHNOLOGY · 2023 COHERENT PHOTONIC COMPUTING ARCHITECTURESUS 11,817,903 · CELESTIAL AI INC. · 2023 NEURAL NETWORK ARCHITECTURE PRUNINGUS 11,663,481 · ADOBE INC. · 2023 Method and System for Bit-Depth Reduction in Artificial Neural NetworksUS 11,106,973 · HONG KONG APPLIED SCIENCE AND TECHNOLOGY RESEARCH INSTITUTE COMPANY, LIMITED. · 2021

Source of record

arXiv