You are designing a training job in an Azure Machine Learning workspace using Automated ML. The compute resource must be able to scale up to a multi-node cluster automatically to handle larger datasets.

Which Azure Machine Learning compute target should you use?