Silicon Photonic 2.5D Interposer Networks for Overcoming Communication Bottlenecks in Scale-out Machine Learning Hardware Accelerators
Fuente:
arXiv
Enregistré dans:
| Auteurs principaux: | , , , |
|---|---|
| Format: | Preprint |
| Publié: |
2024
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866910356202323968 |
|---|---|
| author | Sunny, Febin Taheri, Ebadollah Nikdast, Mahdi Pasricha, Sudeep |
| author_facet | Sunny, Febin Taheri, Ebadollah Nikdast, Mahdi Pasricha, Sudeep |
| contents | Modern machine learning (ML) applications are becoming increasingly complex and monolithic (single chip) accelerator architectures cannot keep up with their energy efficiency and throughput demands. Even though modern digital electronic accelerators are gradually adopting 2.5D architectures with multiple smaller chiplets to improve scalability, they face fundamental limitations due to a reliance on slow metallic interconnects. This paper outlines how optical communication and computation can be leveraged in 2.5D platforms to realize energy-efficient and high throughput 2.5D ML accelerator architectures. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2403_04189 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Silicon Photonic 2.5D Interposer Networks for Overcoming Communication Bottlenecks in Scale-out Machine Learning Hardware Accelerators Sunny, Febin Taheri, Ebadollah Nikdast, Mahdi Pasricha, Sudeep Hardware Architecture Machine Learning Signal Processing Modern machine learning (ML) applications are becoming increasingly complex and monolithic (single chip) accelerator architectures cannot keep up with their energy efficiency and throughput demands. Even though modern digital electronic accelerators are gradually adopting 2.5D architectures with multiple smaller chiplets to improve scalability, they face fundamental limitations due to a reliance on slow metallic interconnects. This paper outlines how optical communication and computation can be leveraged in 2.5D platforms to realize energy-efficient and high throughput 2.5D ML accelerator architectures. |
| title | Silicon Photonic 2.5D Interposer Networks for Overcoming Communication Bottlenecks in Scale-out Machine Learning Hardware Accelerators |
| topic | Hardware Architecture Machine Learning Signal Processing |
| url | https://arxiv.org/abs/2403.04189 |