Amazon Redshift Unleashes Graviton-Powered RG Instances: 2.2x Speed, 30% Cost Cut for Data Warehouses and Lakes
Breaking News: Amazon Redshift Launches Graviton-Based RG Instances
Amazon Web Services today announced Amazon Redshift RG instances, a new instance family powered by AWS Graviton processors. The new instances deliver up to 2.2x faster data warehouse performance than current RA3 instances, at a 30% lower price per vCPU, according to the company.
The integrated data lake query engine enables customers to run SQL analytics across both warehouse tables and Amazon S3 data lakes from a single engine. Performance for Apache Iceberg queries is up to 2.4x faster than RA3, and for Apache Parquet up to 1.5x faster.
Key Details
- Instance types: New RG instances (rg.xlarge, rg.4xlarge, and more) replace RA3 equivalents (ra3.xlplus, ra3.4xlarge) with improved vCPU and memory ratios.
- Availability: Immediate launch via AWS Management Console, CLI, or API.
- Use cases: Designed for high query volumes from BI dashboards, ETL pipelines, near-real-time analytics, and AI agent workloads.
“This blend of speed, cost efficiency, and an integrated data lake query engine makes Redshift RG instances well-suited to handle the high query volumes and low-latency requirements of today’s analytics and agentic AI workloads,” said Rahul Pathak, Vice President of Amazon Redshift, in a press release.
Background
Amazon Redshift has evolved since its 2013 launch, moving from dense compute to RA3 and serverless models. Over a decade, data volumes have surged, and organizations increasingly rely on both structured warehouse tables for frequent access and data lakes for cost-effective storage of diverse datasets.
The rise of AI agents has added new pressure: they query data warehouses at scales that dwarf human usage, driving operational costs higher. Amazon Redshift previously improved BI and ETL performance by up to 7x in March 2026, and the new RG instances build on that momentum.
What This Means
The RG instances represent a significant leap for companies managing hybrid analytics workloads. By combining Graviton’s energy efficiency with an integrated query engine, customers can reduce total analytics costs while simplifying operations. The single-engine approach eliminates the need to switch between systems for warehouse and data lake queries.
“We recommend using the AWS Pricing Calculator with your specific workload patterns to estimate savings,” the company stated. Early adopters can expect lower latency for critical queries and reduced infrastructure complexity as AI and human workloads converge.
Comparison: RG vs. RA3
| Current RA3 | Recommended RG | vCPU | Memory (GB) | Primary Use Case |
|---|---|---|---|---|
| ra3.xlplus | rg.xlarge | 4 | 32 | Small cluster departmental analytics |
| ra3.4xlarge | rg.4xlarge | 16 | 128 | Standard production workloads, medium data volumes |
For more details, see the AWS Redshift pricing page.
Related Articles
- Docker Enables Fully Private AI Image Generation on Local Machines
- Understanding Kubernetes v1.36's Pod-Level Resource Managers: Key Questions Answered
- Kubernetes v1.36 Finalizes Fine-Grained Kubelet Authorization, Closing Critical Security Hole
- Mastering Cloud Cost Optimization: Essential Principles for Modern AI Workloads
- Enhancing CPU Priority in Kubernetes: A New cgroup v1 to v2 Conversion Formula
- Build Your Private Image Generator: Docker Model Runner & Open WebUI Step-by-Step
- Scaling Kubernetes Controllers with Server-Side Sharded Watches
- Cloudflare's AI-Driven Restructuring: A New Blueprint for the Future