What feature provides automatic scaling of compute resources in Snowflake?

Multi-cluster Warehouses are designed to provide automatic scaling of compute resources within Snowflake. This feature is particularly beneficial for handling variable workloads efficiently. When demand increases, additional clusters can be automatically provisioned to handle the additional load, ensuring that performance remains consistent and responsive. Conversely, when the demand decreases, Snowflake can downscale the number of clusters to optimize resource usage and costs.

This automatic scaling capability in Multi-cluster Warehouses allows organizations to achieve near-infinite concurrency and performance consistency without manual intervention, making it a pivotal feature for managing varying levels of data processing demands efficiently.

Other options like Automatic Clustering, Snowflake Tasks, and Persistent Data Caching serve different purposes. Automatic Clustering optimizes the organization of data within the warehouse for performance but does not scale compute resources. Snowflake Tasks automate SQL execution based on a schedule or specific events, while Persistent Data Caching allows for quicker retrieval of frequently accessed data, but neither of these options provides the auto-scaling capabilities associated with compute resources.