NetSmith: An Optimization Framework for Machine-Discovered Network Topologies
Fuente:
arXiv
Guardado en:
| Autores principales: | , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2024
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866913296953638912 |
|---|---|
| author | Green, Conor Thottethodi, Mithuna |
| author_facet | Green, Conor Thottethodi, Mithuna |
| contents | Over the past few decades, network topology design for general purpose, shared memory multicores has been primarily driven by human experts who use their insights to arrive at network designs that balance the competing goals of performance requirements (e.g., latency, bandwidth) and cost constraints (e.g., router radix, router counts). On the other hand, there have been automatic NoC synthesis methods for SoCs to optimize for application-specific communication and objectives such as resource usage or power. Unfortunately, these techniques do not lend themselves to the general-purpose context, where directly applying these previous NoC synthesis techniques in the general-purpose context yields poor results, even worse than expert-designed networks. We design and develop an automatic network design methodology - NetSmith - to design networks for general-purpose, shared memory multicores that comprehensively outperform expert-designed networks.
We employ NetSmith in the context of interposer networks for chiplet-based systems where there has been significant recent work on network topology design (e.g., Kite, Butter Donut, Double Butterfly). NetSmith generated topologies are capable of achieving significantly higher throughput (50% to 75% higher) while also reducing average hop count by 8% to 13.5%) than previous expert-designed and synthesized networks. Full system simulations using PARSEC benchmarks demonstrate that the improved network performance translates to improved application performance with up to 11% mean speedup over previous NoI topologies. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2404_02357 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | NetSmith: An Optimization Framework for Machine-Discovered Network Topologies Green, Conor Thottethodi, Mithuna Hardware Architecture Distributed, Parallel, and Cluster Computing Over the past few decades, network topology design for general purpose, shared memory multicores has been primarily driven by human experts who use their insights to arrive at network designs that balance the competing goals of performance requirements (e.g., latency, bandwidth) and cost constraints (e.g., router radix, router counts). On the other hand, there have been automatic NoC synthesis methods for SoCs to optimize for application-specific communication and objectives such as resource usage or power. Unfortunately, these techniques do not lend themselves to the general-purpose context, where directly applying these previous NoC synthesis techniques in the general-purpose context yields poor results, even worse than expert-designed networks. We design and develop an automatic network design methodology - NetSmith - to design networks for general-purpose, shared memory multicores that comprehensively outperform expert-designed networks. We employ NetSmith in the context of interposer networks for chiplet-based systems where there has been significant recent work on network topology design (e.g., Kite, Butter Donut, Double Butterfly). NetSmith generated topologies are capable of achieving significantly higher throughput (50% to 75% higher) while also reducing average hop count by 8% to 13.5%) than previous expert-designed and synthesized networks. Full system simulations using PARSEC benchmarks demonstrate that the improved network performance translates to improved application performance with up to 11% mean speedup over previous NoI topologies. |
| title | NetSmith: An Optimization Framework for Machine-Discovered Network Topologies |
| topic | Hardware Architecture Distributed, Parallel, and Cluster Computing |
| url | https://arxiv.org/abs/2404.02357 |