AI-300: Operationalizing Machine Learning and Generative AI SolutionsDemo

Question

An organization is deploying generative AI solutions by using Microsoft Foundry to support multiple production workloads.

The organization has these workload requirements:

One workload must be real-time, latency-sensitive, and have predictable global usage patterns that require consistent performance.
One workload must have variable performance and be optimized for cost-efficient operation.

You need to choose a global deployment type for each workload.

Each deployment type may be used once, more than once, or not at all.

Drag & Drop

Global Provisioned

Global Standard

Real-time, predictable

Flexible, cost efficient

Explanation

Global Provisioned uses reserved capacity to deliver predictable throughput and lower latency variance, making it appropriate for real-time, latency-sensitive workloads with consistent demand. Global Standard uses pay-per-token billing and is suited to variable or bursty traffic, which makes it the more cost-efficient option when demand varies.

Learn more

Deployment types for Microsoft Foundry Models

AI-300: Operationalizing Machine Learning and Generative AI SolutionsDemo

AstroTutor

Notes

Readiness

QuestionQ19

Community Discussion