# Mirage: An RNS-Based Photonic Accelerator for DNN Training

arXiv: https://arxiv.org/abs/2311.17323 · HTML: https://mlsanity.com/paper/2311.17323/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- Builds on 20 of its 70 references that are in the map.
- First posted on arXiv 29 Nov 2023.

## Builds on (most cited first, 20 of 20)

- [Attention Is All You Need](https://mlsanity.com/paper/1706.03762/) (2017)
- [Quantization and Training of Neural Networks for Efficient Integer-Arithmetic-Only Inference](https://mlsanity.com/paper/1712.05877/) (2017)
- [YOLO9000: Better, Faster, Stronger](https://mlsanity.com/paper/1612.08242/) (2016)
- [DoReFa-Net: Training Low Bitwidth Convolutional Neural Networks with Low Bitwidth Gradients](https://mlsanity.com/paper/1606.06160/) (2016)
- [Quantizing deep convolutional networks for efficient inference: A whitepaper](https://mlsanity.com/paper/1806.08342/) (2018)
- [Quantized Neural Networks: Training Neural Networks with Low Precision Weights and Activations](https://mlsanity.com/paper/1609.07061/) (2016)
- [Deep Learning with Limited Numerical Precision](https://mlsanity.com/paper/1502.02551/) (2015)
- [Integer Quantization for Deep Learning Inference: Principles and Empirical Evaluation](https://mlsanity.com/paper/2004.09602/) (2020)
- [Scalable Methods for 8-bit Training of Neural Networks](https://mlsanity.com/paper/1805.11046/) (2018)
- [Training of photonic neural networks through in situ backpropagation](https://mlsanity.com/paper/1805.09943/) (2018)
- [Experimentally realized in situ backpropagation for deep learning in nanophotonic neural networks](https://mlsanity.com/paper/2205.08501/) (2022)
- [Training DNNs with Hybrid Block Floating Point](https://mlsanity.com/paper/1804.01526/) (2018)
- [FAST: DNN Training Under Variable Precision Block Floating Point with Stochastic Rounding](https://mlsanity.com/paper/2110.15456/) (2021)
- [CrossLight: A Cross-Layer Optimized Silicon Photonic Neural Network Accelerator](https://mlsanity.com/paper/2102.06960/) (2021)
- [Silicon Photonic Architecture for Training Deep Neural Networks with Direct Feedback Alignment](https://mlsanity.com/paper/2111.06862/) (2021)
- [An Electro-Photonic System for Accelerating Deep Neural Networks](https://mlsanity.com/paper/2109.01126/) (2021)
- [Dynamic Precision Analog Computing for Neural Networks](https://mlsanity.com/paper/2102.06365/) (2021)
- [PCNNA: A Photonic Convolutional Neural Network Accelerator](https://mlsanity.com/paper/1807.08792/) (2018)
- [Adaptive Block Floating-Point for Analog Deep Learning Hardware](https://mlsanity.com/paper/2205.06287/) (2022)
- [A Blueprint for Precise and Fault-Tolerant Analog Neural Networks](https://mlsanity.com/paper/2309.10759/) (2023)

## Built on by (most cited first, 2 of 2)

- [SimPhony: A Device-Circuit-Architecture Cross-Layer Modeling and Simulation Framework for Heterogeneous Electronic-Photonic AI System](https://mlsanity.com/paper/2411.13715/) (2024)
- [Harnessing Photonics for Machine Intelligence](https://mlsanity.com/paper/2604.10841/) (2026)

## Abstract (arXiv)

Photonic computing is a compelling avenue for performing highly efficient matrix multiplication, a crucial operation in Deep Neural Networks (DNNs). While this method has shown great success in DNN inference, meeting the high precision demands of DNN training proves challenging due to the precision limitations imposed by costly data converters and the analog noise inherent in photonic hardware. This paper proposes Mirage, a photonic DNN training accelerator that overcomes the precision challenges in photonic hardware using the Residue Number System (RNS). RNS is a numeral system based on modular arithmetic, allowing us to perform high-precision operations via multiple low-precision modular operations. In this work, we present a novel micro-architecture and dataflow for an RNS-based photonic tensor core performing modular arithmetic in the analog domain. By combining RNS and photonics, Mirage provides high energy efficiency without compromising precision and can successfully train state-of-the-art DNNs achieving accuracy comparable to FP32 training. Our study shows that on average across several DNNs when compared to systolic arrays, Mirage achieves more than 23.8× faster training and 32.1× lower EDP in an iso-energy scenario and consumes 42.8× lower power with comparable or better EDP in an iso-area scenario.