How NVIDIA Scales AI? | Fundamentals of AI Infrastructure
Автор: Abdul Jabbar
Загружено: 2026-01-27
Просмотров: 41
Описание:
In Part 2 of this NVIDIA AI Infrastructure Fundamentals series, we dive deep into how AI scales - from a single server to massive, industrial-scale AI factories.
If you’ve already watched Part 1, where we covered the foundations of AI infrastructure, this video builds directly on that knowledge and explains how NVIDIA scales AI workloads efficiently and reliably.
🔍 What you’ll learn in this video:
0:00 Introduction
2:18 Why do we need to scale AI?
2:57 How to scale AI?
3:36 How scaling works?
4:30 Vertical scaling limits
4:58 How horizontal scaling works
6:00 Network layer services - Infiniband, Ethernet X, ConnectX Nics
7:00 Systems Layer - DGX, HGX, MGX, SuperPOD
12:15 What is NVIDIA Vera Rubin Platform?
18:39 Systems Software Layer - CUDA, Libraries & Frameworks
23:27 Training vs Inference - NEMO, NIMS, TensorRT
Related Video:
Nvidia AI Infrastructure Fundamentals: • What are the building blocks of AI? | NVID...
Повторяем попытку...
Доступные форматы для скачивания:
Скачать видео
-
Информация по загрузке:
