How to Attach a Windows GPU Machine as Compute in Azure ML Studio?

Question

I'm working with a Windows machine with GPU which has both Windows Server 2022 and Azure ARC service, and I would like to add it as an attached compute resource in Azure ML Studio. However, I haven't been able to find any relevant tutorials or documentation on how to achieve this.

Could anyone provide guidance or share any tutorials/documentation that outline the process for attaching a Windows Server GPU machine to Azure ML Studio? Any help is greatly appreciated!

Answer 1

Ramasamy Seenivasagan Greetings!

Could anyone provide guidance or share any tutorials/documentation that outline the process for attaching a Windows Server GPU machine to Azure ML Studio?

To use VMs created outside the Azure Machine Learning workspace, you must first attach them to your workspace. Also, Some VM series aren't supported in Azure Machine Learning.

Did you see the below documentations already? If not, please check and let me know if that helps.

Manage compute resources for model training and deployment in studio

What are compute targets in Azure Machine Learning