Required minimal compute instances to play with Azure ML Designer? RRS feed

  • Question

  • My organization has paid for a few $100 of credit. We wish to learn more about Azure Machine Learning (ML) Designer (formerly ML Studio). Do we need to set up a virtual machine or other compute instance? Does we have to commit to a 1y cost (for which we pay even if we are not using the VM)? Or can we use ML Designer using pay-as-you-go? What is the most cost effective way to use our existing credit to get up & running? Many thanks for any advice. 
    Friday, May 8, 2020 4:06 PM

All replies

  • Hello,

    To use Azure ML Designer local compute cannot be used to run the designer experiment or pipeline. You need to attach either a ML compute cluster or compute instance. In most cases for learning purposes a simple compute instance with minimal compute power should help you run the pipeline or experiment and the instance type/price  depends on the region who are using which is billed accordingly. If you do not plan to use any of the compute instances please ensure they are stopped or deleted as your ML workspace uses blob storage to store details of your experiment and compute is only used to run the experiment when required.

    We hope this helps!!

    If you found this post helpful, please give it a "Helpful" vote. 
    Please remember to mark the replies as answers if they help. 

    Monday, May 11, 2020 5:13 PM