Timezone: »
Silicon-photonics-based optical neural network (ONN) is a promising hardware platform that could represent a paradigm shift in efficient AI with its CMOS-compatibility, flexibility, ultra-low execution latency, and high energy efficiency. In-situ training on the online programmable photonic chips is appealing but still encounters challenging issues in on-chip implementability, scalability, and efficiency. In this work, we propose a closed-loop ONN on-chip learning framework L2ight to enable scalable ONN mapping and efficient in-situ learning. L2ight adopts a three-stage learning flow that first calibrates the complicated photonic circuit states under challenging physical constraints, then performs photonic core mapping via combined analytical solving and zeroth-order optimization. A subspace learning procedure with multi-level sparsity is integrated into L2ight to enable in-situ gradient evaluation and fast adaptation, unleashing the power of optics for real on-chip intelligence. Extensive experiments demonstrate our proposed L2ight outperforms prior ONN training protocols with 3-order-of-magnitude higher scalability and over 30x better efficiency, when benchmarked on various models and learning tasks. This synergistic framework is the first scalable on-chip learning solution that pushes this emerging field from intractable to scalable and further to efficient for next-generation self-learnable photonic neural chips. From a co-design perspective, L2ight also provides essential insights for hardware-restricted unitary subspace optimization and efficient sparse training. We open-source our framework at the link.
Author Information
Jiaqi Gu (The University of Texas at Austin)
Hanqing Zhu (University of Texas, Austin)
Chenghao Feng (University of Texas, Austin)
Zixuan Jiang (University of Texas, Austin)
Ray Chen (University of Texas, Austin)
David Pan (University of Texas, Austin)
More from the Same Authors
-
2022 : An Adversarial Active Sampling-based Data Augmentation Framework for Manufacturable Chip Design »
Mingjie Liu · Haoyu Yang · David Pan · Brucek Khailany · Mark Ren -
2022 : HEAT: Hardware-Efficient Automatic Tensor Decomposition for Transformer Compression »
Jiaqi Gu · Ben Keller · Jean Kossaifi · Anima Anandkumar · Brucek Khailany · David Pan -
2023 Poster: Pre-RMSNorm and Pre-CRMSNorm Transformers: Equivalent and Efficient Pre-LN Transformers »
Zixuan Jiang · Jiaqi Gu · Hanqing Zhu · David Pan -
2022 Spotlight: NeurOLight: A Physics-Agnostic Neural Operator Enabling Parametric Photonic Device Simulation »
Jiaqi Gu · Zhengqi Gao · Chenghao Feng · Hanqing Zhu · Ray Chen · Duane Boning · David Pan -
2022 : HEAT: Hardware-Efficient Automatic Tensor Decomposition for Transformer Compression »
Jiaqi Gu · Ben Keller · Jean Kossaifi · Anima Anandkumar · Brucek Khailany · David Pan -
2022 Poster: NeurOLight: A Physics-Agnostic Neural Operator Enabling Parametric Photonic Device Simulation »
Jiaqi Gu · Zhengqi Gao · Chenghao Feng · Hanqing Zhu · Ray Chen · Duane Boning · David Pan