mlsanityData to 7 Oct 2026

CrossLight: A Cross-Layer Optimized Silicon Photonic Neural Network Accelerator

arXiv 2102.06960 ↗First posted 13 Feb 2021cs.LG cs.AR cs.ETcited by 29

Not on the map. Why?

Abstract (arXiv)

Domain-specific neural network accelerators have seen growing interest in recent years due to their improved energy efficiency and inference performance compared to CPUs and GPUs. In this paper, we propose a novel cross-layer optimized neural network accelerator called CrossLight that leverages silicon photonics. CrossLight includes device-level engineering for resilience to process variations and thermal crosstalk, circuit-level tuning enhancements for inference latency reduction, and architecture-level optimization to enable higher resolution, better energy-efficiency, and improved throughput. On average, CrossLight offers 9.5x lower energy-per-bit and 15.9x higher performance-per-watt at 16-bit resolution than state-of-the-art photonic deep learning accelerators.

Key facts. Cited by 29 papers in the map (data through 7 Oct 2026). Builds on 3 of its 35 references that are in the map. First posted on arXiv 13 Feb 2021.

Builds on

3 of 35 references.

References · 3

  1. 2017In-Datacenter Performance Analysis of a Tensor Processing Unitcited by 610
  2. 2016Neuromorphic Silicon Photonic Networkscited by 66
  3. 2019Digital Electronics and Analog Photonics for Convolutional Neural Networks (DEAP-CNNs)cited by 22

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2021: 2; 2022: 5; 2023: 7; 2024: 8; 2025: 1; 2026: 6.

29 papers.

Sort:

By area:

  1. 2022NeurOLight: A Physics-Agnostic Neural Operator Enabling Parametric Photonic Device Simulationcited by 17
  2. 2022RecLight: A Recurrent Neural Network Accelerator with Integrated Silicon Photonicscited by 12
  3. 2021SONIC: A Sparse Neural Network Inference Accelerator with Silicon Photonics for Energy-Efficient Deep Learningcited by 10
  4. 2022A Silicon Photonic Accelerator for Convolutional Neural Networks with Heterogeneous Quantizationcited by 9
  5. 2023GHOST: A Graph Neural Network Accelerator using Silicon Photonicscited by 7
  6. 2021O-HAS: Optical Hardware Accelerator Search for Boosting Both Acceleration Performance and Development Speedcited by 6
  7. 2022Photonic Reconfigurable Accelerators for Efficient Inference of CNNs with Mixed-Sized Tensorscited by 5
  8. 2023Machine Learning Accelerators in 2.5D Chiplet Platforms with Silicon Photonicscited by 5
  9. 2023M3ICRO: Machine Learning-Enabled Compact Photonic Tensor Core based on PRogrammable Multi-Operand Multimode Interferencecited by 5
  10. 2023SCONNA: A Stochastic Computing Based Optical Accelerator for Ultra-Fast, Energy-Efficient Inference of Integer-Quantized CNNscited by 4
  11. 2024SafeLight: Enhancing Security in Optical Convolutional Neural Network Acceleratorscited by 4
  12. 2023Cross-Layer Design for AI Acceleration with Non-Coherent Optical Computingcited by 3
17 more
  1. 2023Mirage: An RNS-Based Photonic Accelerator for DNN Trainingcited by 2
  2. 2022PhotoFourier: A Photonic Joint Transform Correlator-Based Neural Network Acceleratorcited by 1
  3. 2024Accelerating Neural Networks for Large Language Models and Graph Processing with Silicon Photonicscited by 1
  4. 2024DOCTOR: Dynamic On-Chip Temporal Variation Remediation Toward Self-Corrected Photonic Tensor Acceleratorscited by 1
  5. 2024Silicon Photonic 2.5D Interposer Networks for Overcoming Communication Bottlenecks in Scale-out Machine Learning Hardware Acceleratorscited by 1
  6. 2024OPIMA: Optical Processing-In-Memory for Convolutional Neural Network Accelerationcited by 1
  7. 2025PhotoGAN: Generative Adversarial Neural Network Acceleration with Silicon Photonicscited by 1
  8. 2023Towards Efficient Hyperdimensional Computing Using Photonicscited by 0
  9. 2024A Comparative Analysis of Microrings Based Incoherent Photonic GEMM Acceleratorscited by 0
  10. 2024HEANA: A Hybrid Time-Amplitude Analog Optical Accelerator with Flexible Dataflows for Energy-Efficient CNN Inferencecited by 0
  11. 2024Optical Computing for Deep Neural Network Acceleration: Foundations, Recent Developments, and Emerging Directionscited by 0
  12. 2026PRISM: Breaking the O(n) Memory Wall in Long-Context LLM Inference via O(1) Photonic Block Selectioncited by 0
  13. 2026Harnessing Photonics for Machine Intelligencecited by 0
  14. 2026ROSA: Robust and Energy-Efficient Microring-Based Optical Neural Networks via Optical Shift-and-Add and Layer-Wise Hybrid Mappingcited by 0
  15. 2026DxPTA: An Architecture Design Space Exploration with Optical Dataflow-guided Strategy for HW/SW Co-Design of Photonic Transformer Acceleratorscited by 0
  16. 2026MDTransformer: A Hardware-Software Co-Design of Mode-Division Photonic Transformer Accelerator with Inverse-Designed Coherent Crossbarcited by 0
  17. 2026Hardware Trojan Threats to Multi-Chiplet Photonic Neural Network Acceleratorscited by 0