Start leveraging powerful machine
learning infrastructure now with
Amazon SageMaker
The easiest and fastest way to benefit from powerful machine learning infrastructure on AWS is to deploy
Amazon SageMaker. This fully managed service brings together a broad set of essential capabilities,
such as data labeling, data preparation, feature engineering, statistical bias detection, AutoML, training,
tuning, hosting, explainability, monitoring, and workflows.
NVIDIA GPUs, available on Amazon EC2, can help developers significantly accelerate these stages in
developing machine learning algorithms in Amazon SageMaker, including model training and inference,
reducing the overall costs of building and deploying machine learning applications.
To achieve these benefits, NVIDIA offers the NGC catalog, a comprehensive collection of GPU-optimized
libraries, pre-trained AI models for computer vision, conversational AI and recommenders, application
frameworks, and inference serving solutions like Triton Inference Server to simplify the deployment of
machine learning models on CPUs and GPUs. The software in the NGC catalog is constantly optimized,
allowing developers to easily access the latest NVIDIA innovations. Monthly releases give users access to
the latest features and performance improvements.
The NVIDIA NGC catalog is available directly on the AWS Marketplace, making it seamless for users to
pull and run GPU-optimized containers and models within Amazon SageMaker or other AWS services like
Amazon Elastic Kubernetes Service (Amazon EKS) or Amazon Elastic Container Service (Amazon ECS).
12