Chapter 3: AI Platforms and Ecosystem¶
Modern AI relies on a rich ecosystem of tools, frameworks, and hardware that work together to move data, train models, monitor workflows, and deploy AI applications at scale.
Building a successful AI system is not just about models — it requires the right orchestration, automation, and acceleration layers.
In this chapter, you will:
- Explore key CNCF MLOps toolchains such as Airflow, MLflow, Kubeflow, and Spark, and understand how they streamline data processing, experiment tracking, orchestration, and production deployment
- Learn how NVIDIA’s acceleration stack — including CUDA, TensorRT, cuDNN, NCCL, and DGX systems — provides the compute foundation powering modern AI workloads from training to inference
This chapter connects the software and hardware layers that make today’s high-performance AI systems possible.