mlsanityData to 7 Oct 2026

SONIC: A Sparse Neural Network Inference Accelerator with Silicon Photonics for Energy-Efficient Deep Learning

arXiv 2109.04459 ↗First posted 9 Sep 2021cs.LG cs.AR cs.NEcited by 10

Not on the map. Why?

Abstract (arXiv)

Sparse neural networks can greatly facilitate the deployment of neural networks on resource-constrained platforms as they offer compact model sizes while retaining inference accuracy. Because of the sparsity in parameter matrices, sparse neural networks can, in principle, be exploited in accelerator architectures for improved energy-efficiency and latency. However, to realize these improvements in practice, there is a need to explore sparsity-aware hardware-software co-design. In this paper, we propose a novel silicon photonics-based sparse neural network inference accelerator called SONIC. Our experimental analysis shows that SONIC can achieve up to 5.8x better performance-per-watt and 8.4x lower energy-per-bit than state-of-the-art sparse electronic neural network accelerators; and up to 13.8x better performance-per-watt and 27.6x lower energy-per-bit than the best known photonic neural network accelerators.

Key facts. Cited by 10 papers in the map (data through 7 Oct 2026). Builds on 10 of its 28 references that are in the map. First posted on arXiv 9 Sep 2021.

Builds on

10 of 28 references.

References · 10

  1. 2014Very Deep Convolutional Networks for Large-Scale Image Recognitioncited by 15,267
  2. 2015Deep Compression: Compressing Deep Neural Networks with Pruning, Trained Quantization and Huffman Codingcited by 2,291
  3. 2017To prune, or not to prune: exploring the efficacy of pruning for model compressioncited by 596
  4. 2016Faster CNNs with Direct Sparse Convolutions and Guided Pruningcited by 39
  5. 2017NullHop: A Flexible Convolutional Neural Network Accelerator Based on Sparse Representations of Feature Mapscited by 39
  6. 2021CrossLight: A Cross-Layer Optimized Silicon Photonic Neural Network Acceleratorcited by 29
  7. 2021A Survey on Silicon Photonics for Deep Learningcited by 23
  8. 2021ROBIN: A Robust Optical Binary Neural Network Acceleratorcited by 15
  9. 2020Modeling Silicon-Photonic Neural Networks under Uncertaintiescited by 7
  10. 2021BPLight-CNN: A Photonics-based Backpropagation Accelerator for Deep Learningcited by 4

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2021: 1; 2022: 1; 2023: 3; 2024: 3; 2025: 1; 2026: 1.

10 papers.

  1. 2022RecLight: A Recurrent Neural Network Accelerator with Integrated Silicon Photonicscited by 12
  2. 2023GHOST: A Graph Neural Network Accelerator using Silicon Photonicscited by 7
  3. 2023R-TOSS: A Framework for Real-Time Object Detection using Semi-Structured Pruningcited by 4
  4. 2024SafeLight: Enhancing Security in Optical Convolutional Neural Network Acceleratorscited by 4
  5. 2021Pruning Coherent Integrated Photonic Neural Networks Using the Lottery Ticket Hypothesiscited by 3
  6. 2023Cross-Layer Design for AI Acceleration with Non-Coherent Optical Computingcited by 3
  7. 2024Accelerating Neural Networks for Large Language Models and Graph Processing with Silicon Photonicscited by 1
  8. 2025Designing Object Detection Models for TinyML: Foundations, Comparative Analysis, Challenges, and Emerging Solutionscited by 1
  9. 2024Optical Computing for Deep Neural Network Acceleration: Foundations, Recent Developments, and Emerging Directionscited by 0
  10. 2026Accelerating Diffusion Models for Generative AI Applications with Silicon Photonicscited by 0