Amazon introduces Redshift RG instances powered by AWS Graviton processors, providing up to 2.4x performance improvements on data lake queries and reducing compute costs by 30% per vCPU while streamlining data architecture.
- Up to 2.4x faster combined data warehouse and lake queries vs RA3
- 30% lower price per vCPU with AWS Graviton processors
- Eliminates Spectrum scanning fees, reducing overall analytics expenses
Infrastructure signal
The new RG instance family utilizes AWS Graviton processors to deliver substantial improvements in speed and cost efficiency for Amazon Redshift workloads. Benchmarks show RG instances can perform data warehouse tasks up to 2.2 times faster than prior RA3 instances, while data lake queries, especially those using Apache Iceberg and Parquet formats, achieve up to 2.4 times faster execution.
A key infrastructure innovation is the integrated data lake query engine running directly on the RG cluster nodes. This removes the need for Redshift Spectrum to process external table queries separately and eliminates the associated $5 per terabyte scanning fees. Running all queries within customer VPCs with existing IAM roles improves security and reduces complexity.
Developer impact
Developers and data engineers benefit from simplified workflows as Redshift RG instances unify data warehouse and data lake querying under a single engine. Existing external tables, schemas, and query syntax are fully compatible, removing the need for code or schema changes when migrating from older instances or Spectrum setups.
Improved query latency for BI dashboards, ETL pipelines, and AI-driven workloads enables faster iteration and more responsive analytics applications. Developers can migrate clusters via the AWS console, CLI, or API with automated compatibility and cost-estimation tools, accelerating adoption without disruption.
What teams should watch
Cloud architects and infrastructure teams should evaluate workload patterns to gauge cost savings and performance gains from the RG instances. Leveraging the AWS Pricing Calculator is recommended to model expected reductions in operational expense, especially for high-volume query environments integrating both data lakes and warehouses.
Data platform and analytics teams must monitor query performance and adapt deployment strategies to benefit from the reduced reliance on Redshift Spectrum and the new integrated lake engine. Observability around query throughput and latency will be crucial to assess the real-world impact on SLAs and end-user experience.