Question 25
Domain 3: Deployment and Orchestration of ML WorkflowsA team needs to deploy a machine learning model to devices that run in the field rather than in a centralized cloud environment. Which AWS approach is specifically aligned with edge deployment requirements in this scenario?
Correct answer: B
Explanation
For machine learning workloads that must run on devices outside a centralized cloud environment, use AWS services designed for edge deployment rather than standard cloud-only hosting options. — Edge deployment: SageMaker Edge Manager, AWS IoT Greengrass.
Why each option is right or wrong
A. Use a centralized cloud deployment service intended for workloads running only in the cloud
Edge deployment targets devices in the field, not workloads running only in a centralized cloud environment.
B. Use SageMaker Edge Manager or AWS IoT Greengrass to support deployment on edge devices
The scenario requires deployment to devices that operate in the field. The source material identifies SageMaker Edge Manager and AWS IoT Greengrass as the AWS services for edge deployment, matching that requirement directly.
C. Use a general-purpose storage service because model files can be placed there for devices to access
Storage alone does not identify an AWS edge deployment approach for running models on field devices.
D. Use a relational database service because device metadata can be stored centrally during deployment
Databases may store metadata, but they are not the AWS services identified for edge deployment.