Exam DP-100 topic 3 question 153 discussion

Azure Container Instances (ACI) does support GPU-based deployments, which can be used to accelerate the execution of deep learning models. You can use ACI to deploy a GPU-enabled model as a web service in an Azure Machine Learning workspace. Regarding Azure Machine Learning compute clusters, you are also correct that they cannot be used as a deployment target for web services. Compute clusters are used for training and batch scoring workloads, not for deploying models as web services. In summary, the correct answer to your question is A. Azure Kubernetes Service (AKS) and B. Azure Container Instances (ACI). Check the link: https://learn.microsoft.com/en-us/azure/container-instances/container-instances-gpu SaM

answer is AD.

Answer should be A and D

ACI is for small/insignificant deployments (low CPU and RAM), so by logic it's not making sense to have GPU enabled on them.

To run certain compute-intensive workloads on Azure Container Instances, deploy your container groups with GPU resources. The container instances in the group can access one or more NVIDIA Tesla GPUs while running container workloads such as CUDA and deep learning applications.

Only A&B support GPU and should be used for inference (web service). D can have GPU but meant for training.

this is correct

AB correct

Answer - AD Azure Compute Instances do not support GPU, only use for low-scale CPU-based workloads. Reference https://learn.microsoft.com/en-us/azure/machine-learning/concept-compute-target?view=azureml-api-2

ACI doesn't support GPU

Answer AD