Advanced9h 35m4 modules · 19 leçons0 inscrits

DGX and AI in the NVIDIA Data Center

End-to-end NVIDIA reference architectures: the anatomy of a DGX system, DGX SuperPOD, rack-scale NVLink domains with GB200 NVL72, InfiniBand networking and BlueField DPUs, GPUDirect storage. Then real-world operations: Base Command, NVIDIA AI Enterprise, Slurm and Kubernetes on DGX, MIG, distributed training and resilience, DCGM, sovereignty — all the way to the DGX-Ready Data Center program and DGX Cloud.

Par Équipe Curriculum Nexus

119€

Accès complet · Certificat inclus

S'inscrire maintenant

Programme du cours

4 modules

01
Module
DGX foundations and infrastructure
The physical anatomy of an NVIDIA AI data center: from the DGX node to the SuperPOD, InfiniBand networking, and parallel storage.
5 leçons·2h 30

Orchestration and multi-node operations

Running the cluster: Base Command, Slurm, Kubernetes + GPU Operator, and managing multi-user GPU queues.

5 leçons·2h 30

DGX and AI in the NVIDIA Data Center

Programme du cours

DGX foundations and infrastructure

Orchestration and multi-node operations

Large-scale training

Operating the data center