Designing scalable AI clusters: architecture and networking for multi-site deployments
Автор: RCR Wireless News
Загружено: 2025-12-09
Просмотров: 38
Описание:
Scaling AI workloads across multiple data centers requires more than raw compute—it demands intelligent cluster design and high-performance networking. This session explores strategies for building efficient, multi-node AI clusters, placing accelerators and CPUs/XPUs optimally, and designing resilient chip-to-chip and optical fiber networks that deliver low-latency, high-throughput performance while minimizing energy use.
Speakers
Raymond Chik, Vice Chair IEEE AI Hardware & Infrastructure Working Group, IEEE AI Coalition
Clayton Wagar, Principal Consulting Engineer for AI and HPC, Nokia
Susana Schwartz, Technology Editor, RCRTech
Повторяем попытку...
Доступные форматы для скачивания:
Скачать видео
-
Информация по загрузке:
