NVIDIA Webinar
Build AI Factories with Scalable Data Centers and High Performance Networking
AI factories are redefining data centers and enabling the next era of AI. Unlike traditional data centers built for general-purpose workloads, AI factories are purpose-built to extract value from AI. They orchestrate the full AI lifecycle from data ingestion to training, fine-tuning, and high-volume inference at scale.
As AI workloads grow in complexity, they demand robust infrastructure across every layer, from power and cooling to high-performance networking and system orchestration.
In this webinar series, NVIDIA experts will walk through the essential building blocks behind modern AI factories, covering both physical infrastructure and reference network architecture - each critical to supporting large-scale, multi-GPU systems for AI training and inference.
Webinar 1: Enable AI Factories with Optimized Data Center Infrastructure
This webinar focuses on the physical infrastructure required to deploy rack-scale compute systems, including floor planning, liquid cooling design, power delivery, and integration with building management systems - using NVIDIA NVL72 as a reference.
Webinar 2: Build AI Factories with NVIDIA RA and High-Performance Networking
This webinar introduces NVIDIA’s Cloud Partner Reference Architecture (NCP RA) for scalable AI factories, covering key components such as east-west networking, converged Ethernet for storage I/O, and in-band management. We’ll also cover rail-optimized topologies, repeatable building blocks that scale to tens of thousands of nodes, and control plane strategies to support multi-tenancy.
Learnings
In webinar 1, you'll learn:
- Optimized layouts with modular compute and network zones
- Liquid cooling, Power delivery and redundancy strategies
- Integrated Building Management System (BMS) - Understand the importance of RA topologies in ensuring high performance, efficient operations, and availability
In webinar 2, you'll learn:
- Networking technologies in NVIDIA Cloud Partner Reference Architecture (NCP RA)
- Key components, compute, networks, storage and management in NCP RA
- Understand the importance of RA topologies in ensuring high performance, efficient operations, and availability
