Amazon is launching RG instances for Redshift, powered by AWS Graviton processors. The new instance family offers up to 2.2 times better performance for data warehouse workloads than the current RA3 instances, at a 30 percent lower price per vCPU.
The RG instances feature an integrated data lake query engine that performs SQL analyses on both warehouse tables and data in Amazon S3. Until now, Redshift relied on Redshift Spectrum for data lake queries, which charged $5 per terabyte processed. With the new instances, those costs disappear entirely: queries remain within the VPC boundary and use existing IAM roles.
For Apache Iceberg, the integrated engine delivers up to 2.4x performance improvement over RA3. For Apache Parquet, that improvement is up to 1.5 times. The query engine is enabled by default on new clusters, without requiring any changes to application code.
Migration and Availability
Users can migrate existing RA3 clusters to RG instances in two ways: via Elastic Resize (an in-place migration with 10 to 15 minutes of downtime) or via Snapshot and Restore. Existing external tables, schemas, and query syntax, including existing Spectrum queries, remain fully intact.
RG instances are available immediately in multiple AWS regions, including Europe (Frankfurt, Ireland, London, Paris, and Stockholm). Users can choose between On-Demand Instances with hourly billing and Reserved Instances for lower long-term costs.