Usage Metering - Amazon SageMaker

Usage Metering

The costs incurred for running Amazon SageMaker Studio notebooks, interactive shells, consoles, and terminals are based on Amazon Elastic Compute Cloud (Amazon EC2) instance usage.

When you launch the following resources, you must choose a SageMaker image and kernel:

From the Studio Launcher

  • Notebook

  • Interactive Shell

  • Image Terminal

From the File menu

  • Notebook

  • Console

When launched, the resource is run on an Amazon EC2 instance of an instance type based on the chosen SageMaker image and kernel. If an instance of that type was previously launched and is available, the resource is run on that instance. If an instance of that type is not available, the resource is run on a new default instance of that type.

For CPU based images, the default instance type is ml.t3.medium. For GPU based images, the default instance type is ml.g4dn.xlarge.

The costs incurred are based on the instance type and the number of instances of each instance type. You are billed separately for each instance.

Metering starts when an instance is created. Metering ends when the SageMaker image running on the instance is shut down. For information on how to shut down a SageMaker image, see Shut Down Resources.

Important

You must shut down the SageMaker image on the instance to stop incurring charges. If you shut down the notebook running on the instance but don't shut down the SageMaker image, you will still incur charges.

When you open multiple notebooks on the same instance type, the notebooks run on the same instance even if they're using different kernels. You are billed only for the time that one instance is running.

You can change the instance type from within the notebook after you open it. For more information, see Change an Instance Type.

Metering Example

In this example, the customer performed the following actions:

  • Opened notebook #1. Studio opens this on a ml.t3.medium instance type.

  • After an hour, opened notebook #2. This notebook runs on the same instance as notebook #1, a ml.t3.medium instance type.

  • After another hour, opened notebook #3 and immediately switched to a ml.m5.large instance type.

  • After another hour, shut down the kernels on all instance types.

The notebooks ran for the following amount of time:

  • Notebook #1: 3 hours

  • Notebook #2: 2 hours

  • Notebook #3: 1 hours

The customer is billed for the following usage:

3 hours at the ml.t3.medium rate + 1 hour at the ml.m5.large rate

During the second and third hours, notebook #1 and notebook #2 ran on the same ml.t3.medium instance.

Important

This topic discusses costs only as it applies to the amount of time an instance is running. Additional costs can be incurred, for example, from running the code in a notebook, such as compute time due to a training job.

For more information about billing along with pricing examples, see Amazon SageMaker Pricing. On the pricing page, under one of the AWS Regions supported by Amazon SageMaker Studio, review the Building section where two types of notebooks are listed: On-Demand ML Notebook Instances and Amazon SageMaker Studio Notebook Instances. In this example, we're discussing Amazon SageMaker Studio Notebook Instances.