A Highly Scalable LLM Clusters with Optical Interconnect
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , , , , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2024
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866918151078281216 |
|---|---|
| author | Han, Xinchi Lv, Yongxi Jiang, Weihao Zhang, Shuyuan Mao, Yingming Zhao, Shizhen Liu, ZhuoRan Liu, Zhuotao Cao, Peirui Liu, Ximeng Wang, Xinbing |
| author_facet | Han, Xinchi Lv, Yongxi Jiang, Weihao Zhang, Shuyuan Mao, Yingming Zhao, Shizhen Liu, ZhuoRan Liu, Zhuotao Cao, Peirui Liu, Ximeng Wang, Xinbing |
| contents | Recent years have witnessed the adoption of optical circuit switch (OCS) technology. How to design the physical topology, defined by the physical wiring between electrical switching equipments and the OCS, is fundamental to designing efficient OCS-based clusters. We identify three features to evaluate the quality of a physical topology design: logical topology compatibility, cluster scalability, and topology engineering polynomial-solvability. However, none of existing physical topologies has achieved these three features simultaneously. This paper explores designing an optimal physical topology that simultaneously maximizes all. We begin by analyzing the importance of these features in OCS-based cluster and examine the limitations of current designs. Leveraging a proposed \emph{Symmetric Integer Matrix Decomposition Theorem}, we outline a general approach for designing optimal physical topologies and introduce \textbf{Cross Wiring} as a concrete instantiation. The feasibility and advantages of Cross Wiring are verified through a 128-NPU testbed and large-scale real-trace-based simulations. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2411_01503 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | A Highly Scalable LLM Clusters with Optical Interconnect Han, Xinchi Lv, Yongxi Jiang, Weihao Zhang, Shuyuan Mao, Yingming Zhao, Shizhen Liu, ZhuoRan Liu, Zhuotao Cao, Peirui Liu, Ximeng Wang, Xinbing Networking and Internet Architecture Recent years have witnessed the adoption of optical circuit switch (OCS) technology. How to design the physical topology, defined by the physical wiring between electrical switching equipments and the OCS, is fundamental to designing efficient OCS-based clusters. We identify three features to evaluate the quality of a physical topology design: logical topology compatibility, cluster scalability, and topology engineering polynomial-solvability. However, none of existing physical topologies has achieved these three features simultaneously. This paper explores designing an optimal physical topology that simultaneously maximizes all. We begin by analyzing the importance of these features in OCS-based cluster and examine the limitations of current designs. Leveraging a proposed \emph{Symmetric Integer Matrix Decomposition Theorem}, we outline a general approach for designing optimal physical topologies and introduce \textbf{Cross Wiring} as a concrete instantiation. The feasibility and advantages of Cross Wiring are verified through a 128-NPU testbed and large-scale real-trace-based simulations. |
| title | A Highly Scalable LLM Clusters with Optical Interconnect |
| topic | Networking and Internet Architecture |
| url | https://arxiv.org/abs/2411.01503 |