Amazon Redshift Serverless introduces AI-driven scaling as default for new workgroups

Amazon Redshift Serverless has implemented AI-driven scaling and optimization as the default setting for all new workgroups. This advanced feature employs machine learning to anticipate computing needs and automatically adjusts resources before queries are queued, thereby enhancing price-performance without the need for manual adjustments.

The recent update broadens support for workloads with a Base RPU (Redshift Processing Unit) range of 8–512 RPU, a change from the previous range of 32–512 RPU. This adjustment effectively lowers the entry cost for utilizing AI-driven scaling. With this feature, Amazon Redshift continuously monitors workload patterns and dynamically adjusts computing resources based on factors such as query complexity, data volume, and anticipated data scan size.

Users have the flexibility to use a price-performance slider to determine their preference for prioritizing cost, performance, or a balanced approach. Additionally, Amazon Redshift incorporates further optimizations including automatic materialized views and automatic table design optimization to align with the selected targets. Configuration of price-performance targets can be managed through the AWS Management Console or via Amazon Redshift API operations, with the option to modify these targets post-creation of the workgroup.

The AI-driven scaling and optimization feature is available in all AWS Regions where Amazon Redshift Serverless is offered. For further details, users can refer to the Amazon Redshift Serverless product page and the documentation on AI-driven scaling and optimization.