Amazon Redshift introduces AWS Graviton-based RG instances with an integrated data lake query engine

Last month, AWS announced a significant upgrade to Amazon Redshift with the introduction of RG instances powered by AWS Graviton processors. If you’ve been working with Redshift’s RA3 instances, this matters to you: RG instances deliver up to 2.4x faster performance on the same workloads while costing 30% less per vCPU. But the performance bump isn’t the only story here—the integrated data lake query engine fundamentally changes how you can structure your analytics infrastructure.

Here’s what’s actually happening under the hood. AWS Graviton is a custom-built ARM processor that AWS designs specifically for their workloads. Unlike traditional x86 chips, Graviton is optimized for price-to-performance at scale, which translates to real savings in infrastructure costs. More importantly for data warehouse work, RG instances now include an integrated query engine that can directly query Apache Iceberg tables—an open table format that’s been gaining adoption in the data lake community. This means you’re no longer locked into Redshift’s proprietary format; you can query data sitting in your S3 data lake using open standards, then blend those results with your traditional data warehouse queries. The architecture essentially eliminates the friction between your data lake and data warehouse, which previously required ETL jobs and data movement.

Let’s think about practical scenarios. Imagine you’re running a financial services company with real-time transaction data in your data lake (stored as Iceberg tables) and historical customer analytics in your traditional Redshift cluster. Before RG instances, joining these datasets required extracting data from S3, loading it into Redshift, and managing that pipeline. Now you can query both directly within Redshift in a single query, without the intermediate copy step. For teams managing large-scale analytics—think retail companies analyzing clickstream data alongside inventory records, or healthcare organizations correlating patient records with billing data—this eliminates hours of ETL time and simplifies your data architecture significantly.

The cost-performance equation makes this especially relevant for growing teams operating on cloud budgets. If you’re currently running RA3 instances, a migration to RG instances could reduce your annual spend while actually improving query performance. The move also addresses a broader industry trend: the blurring of lines between data warehouses and data lakes. Organizations are increasingly moving away from proprietary formats and choosing open table formats like Iceberg and Apache Hudi. RG instances put Redshift on the right side of that shift, letting you build analytics infrastructure that isn’t vendor-locked and can evolve with your data platform strategy. If you’re planning infrastructure for 2024, this deserves serious evaluation in your cost and architecture reviews.