Guardado en:
| Autores principales: | , , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2026
|
| Materias: | |
| Acceso en línea: | https://arxiv.org/abs/2605.00032 |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866910182889488384 |
|---|---|
| author | Zhang, Huifan Hu, Yun Sheng, Caizhi Qu, Yurui Zhou, Pingqiang |
| author_facet | Zhang, Huifan Hu, Yun Sheng, Caizhi Qu, Yurui Zhou, Pingqiang |
| contents | This work presents ROSA, a microring-based optical neural network architecture that improves robustness and energy efficiency using an optical shift-and-add (OSA) module and a layer-wise hybrid mapping strategy. It introduces a noise-aware voltage-to-weight model considering DAC and thermal variations, and a workload-aware framework to co-optimize MRR array size and layer-wise dataflow. Optimized arrays reduce the aggregated relative energy-delay product (EDP) by 64% and 26% compared with DEAP-CNNs and a general compact array, respectively. OSA further contributes 29% EDP reduction. The proposed hybrid mapping strategy improves CIFAR-10 accuracy by 8.3% over weight-stationary mapping while achieving an average 54.7% lower EDP than DEAP-CNNs. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2605_00032 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | ROSA: Robust and Energy-Efficient Microring-Based Optical Neural Networks via Optical Shift-and-Add and Layer-Wise Hybrid Mapping Zhang, Huifan Hu, Yun Sheng, Caizhi Qu, Yurui Zhou, Pingqiang Hardware Architecture Machine Learning This work presents ROSA, a microring-based optical neural network architecture that improves robustness and energy efficiency using an optical shift-and-add (OSA) module and a layer-wise hybrid mapping strategy. It introduces a noise-aware voltage-to-weight model considering DAC and thermal variations, and a workload-aware framework to co-optimize MRR array size and layer-wise dataflow. Optimized arrays reduce the aggregated relative energy-delay product (EDP) by 64% and 26% compared with DEAP-CNNs and a general compact array, respectively. OSA further contributes 29% EDP reduction. The proposed hybrid mapping strategy improves CIFAR-10 accuracy by 8.3% over weight-stationary mapping while achieving an average 54.7% lower EDP than DEAP-CNNs. |
| title | ROSA: Robust and Energy-Efficient Microring-Based Optical Neural Networks via Optical Shift-and-Add and Layer-Wise Hybrid Mapping |
| topic | Hardware Architecture Machine Learning |
| url | https://arxiv.org/abs/2605.00032 |