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

2021/12/11 by Sanmitra Banerjee, Mahdi Nikdast, Banerjee, Sanmitra +5 · 1 citation
Computer Science · Engineering · #Emerging Technologies (cs.ET) #FOS: Computer and information sciences #Machine Learning (cs.LG) #Neural Networks and Reservoir Computing #Optical Network Technologies #Photonic and Optical Devices

paper · pdf · doi:10.48550/arxiv.2112.06098

openalex publication_date 2021/12/11 · openalex created_date 2021/12/31 · openalex updated_date 2026/07/28

Abstract

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.

Cited by

• LightPro: A Linear Photonic Processor with Full Programmability

Related

• Pruning Coherent Integrated Photonic Neural Networks Using the Lottery Ticket Hypothesis
• Optimizing Coherent Integrated Photonic Neural Networks under Random\n Uncertainties
• PCNNA: A Photonic Convolutional Neural Network Accelerator
• Massively Scalable Wavelength Diverse Integrated Photonic Linear Neuron
• Control-free and efficient integrated photonic neural networks via hardware-aware training and pruning