A Silicon Photonic Accelerator for Convolutional Neural Networks with Heterogeneous Quantization

2022/05/17 by Febin Sunny, Mahdi Nikdast, Sunny, Febin +3 · 1 citation
Computer Science · Engineering · Physics and Astronomy · #Advanced Fiber Laser Technologies #FOS: Computer and information sciences #Hardware Architecture (cs.AR) #Machine Learning (cs.LG) #Neural Networks and Reservoir Computing #Photonic and Optical Devices

paper · pdf · doi:10.48550/arxiv.2205.11244

openalex publication_date 2022/05/17 · openalex created_date 2022/05/26 · openalex updated_date 2026/07/28

Abstract

Parameter quantization in convolutional neural networks (CNNs) can help generate efficient models with lower memory footprint and computational complexity. But, homogeneous quantization can result in significant degradation of CNN model accuracy. In contrast, heterogeneous quantization represents a promising approach to realize compact, quantized models with higher inference accuracies. In this paper, we propose HQNNA, a CNN accelerator based on non-coherent silicon photonics that can accelerate both homogeneously quantized and heterogeneously quantized CNN models. Our analyses show that HQNNA achieves up to 73.8x better energy-per-bit and 159.5x better throughput-energy efficiency than state-of-the-art photonic CNN accelerators.

Cited by

• Opto-ViT: Architecting a Near-Sensor Region of Interest-Aware Vision Transformer Accelerator with Silicon Photonics

Related

• PCNNA: A Photonic Convolutional Neural Network Accelerator
• A Winograd-based Integrated Photonics Accelerator for Convolutional Neural Networks
• Photonic Reconfigurable Accelerators for Efficient Inference of CNNs with Mixed-Sized Tensors
• SCONNA: A Stochastic Computing Based Optical Accelerator for Ultra-Fast, Energy-Efficient Inference of Integer-Quantized CNNs
• SONIC: A Sparse Neural Network Inference Accelerator with Silicon Photonics for Energy-Efficient Deep Learning