Terminate your instance when you no longer need it and stop incurring charges for that instance.

Monitor scheduled events for your instances.

Add an instance store volume to an Amazon EC2 instance at launch and then prepare it for use.

You can use Amazon EC2 to create your key pairs, or you can use a third-party tool to create your key pairs, and then import them to Amazon EC2.

By default, simplified automatic recovery monitors all supported running instances. In the event that a system status check failure is detected, simplified automatic recovery attempts to remediate the instance to a healthy state. Simplified automatic recovery doesn't operate during service events in the AWS Health Dashboard. For more information, see

Learn about Amazon SageMaker capabilities to build, train, and deploy geospatial models.

Compile deep learning (DL) models to achieve an improved training efficiency by using Amazon SageMaker Training Compiler.

Use Amazon Elastic Inference (EI) to speed up the throughput and decrease latency for inference in Amazon SageMaker

Inference recommendation and load balancing testing with Amazon SageMaker Inference Recommender.

Define least-privilege user permissions for machine learning personas with Amazon SageMaker Role Manager.