You need to isolate training workloads while remaining cost-aware to address Fabrikam Inc.’s issues, constraints, and technical requirements. What should you implement?
Correct Answer: D
Explanation:
Managed compute targets with autoscaling are the best option because they allow training workloads to scale up or down based on demand while optimizing cost. Instead of keeping resources running at full capacity all the time, autoscaling ensures that compute resources are only provisioned when needed and automatically reduced during idle periods. This helps isolate training workloads effectively—since each job can still run in its own managed environment—while also controlling operational costs. Compared to fixed-size or dedicated clusters, this approach provides better flexibility and cost efficiency, making it ideal for environments like Fabrikam Inc. where both isolation and budget awareness are required.
Question 2
You need to standardize how Fabrikam Inc. manages machine learning assets. Which action should you perform first?
Correct Answer: B
Explanation:
A shared Azure Machine Learning workspace is the foundational step for standardizing how machine learning assets are managed across an organization. The workspace acts as a central environment where datasets, models, environments, pipelines, and compute resources are created, organized, and governed. By establishing a shared workspace first, Fabrikam Inc. ensures that all teams are working within a consistent structure, which enables proper collaboration, access control, and lifecycle management of ML assets. Once the workspace is in place, other capabilities—such as registering assets in an Azure Machine Learning registry or deploying managed online endpoints—can be built on top of it. Without a standardized workspace, asset management would remain fragmented and inconsistent, making governance and reuse difficult.
Demo Practice Mode
You are viewing only the questions marked as Demo.