Accelerating Diffusion Models for Generative AI Applications with Silicon Photonics
Fuente:
arXiv
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866912952152489984 |
|---|---|
| author | Suresh, Tharini Afifi, Salma Pasricha, Sudeep |
| author_facet | Suresh, Tharini Afifi, Salma Pasricha, Sudeep |
| contents | Diffusion models have revolutionized generative AI, with their inherent capacity to generate highly realistic state-of-the-art synthetic data. However, these models employ an iterative denoising process over computationally intensive layers such as UNets and attention mechanisms. This results in high inference energy on conventional electronic platforms, and thus, there is an emerging need to accelerate these models in a sustainable manner. To address this challenge, we present a novel silicon photonics-based accelerator for diffusion models. Experimental evaluations demonstrate that our photonic accelerator achieves at least 3x better energy efficiency and 5.5x throughput improvement compared to state-of-the-art diffusion model accelerators. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2603_07626 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Accelerating Diffusion Models for Generative AI Applications with Silicon Photonics Suresh, Tharini Afifi, Salma Pasricha, Sudeep Hardware Architecture Machine Learning Diffusion models have revolutionized generative AI, with their inherent capacity to generate highly realistic state-of-the-art synthetic data. However, these models employ an iterative denoising process over computationally intensive layers such as UNets and attention mechanisms. This results in high inference energy on conventional electronic platforms, and thus, there is an emerging need to accelerate these models in a sustainable manner. To address this challenge, we present a novel silicon photonics-based accelerator for diffusion models. Experimental evaluations demonstrate that our photonic accelerator achieves at least 3x better energy efficiency and 5.5x throughput improvement compared to state-of-the-art diffusion model accelerators. |
| title | Accelerating Diffusion Models for Generative AI Applications with Silicon Photonics |
| topic | Hardware Architecture Machine Learning |
| url | https://arxiv.org/abs/2603.07626 |