Since its launch in 2013, Amazon Redshift has consistently redefined cloud data warehousing by delivering enterprise-grade performance at a fraction of on-premises costs. With each generation—from dense compute to RA3 instances, and from provisioned to serverless—Redshift has made every query cheaper, faster, and more efficient. Now, Amazon Redshift introduces a new chapter with RG instances, built on AWS Graviton processors, combining blazing speed with an integrated data lake query engine to tackle the most demanding analytics and AI workloads.
A New Era of Performance and Cost Efficiency
The RG instance family is designed to deliver a significant leap in performance. Compared to current RA3 instances, RG instances run data warehouse workloads up to 2.2x faster, while offering a 30% lower price per vCPU. This combination of speed and cost savings makes it ideal for organizations looking to scale their analytics without breaking the bank.
How Fast?
Performance improvements are not just theoretical. In real-world scenarios, Redshift RG instances accelerate business intelligence (BI) dashboards, ETL pipelines, and near-real-time analytics. For instance, Amazon Redshift already improved new query performance by up to 7x in March 2026, slashing response times for low-latency SQL queries used by BI tools, dashboards, and autonomous AI agents. With RG instances, that performance is further amplified.
Cost Savings at Scale
The 30% reduction in price per vCPU directly lowers total cost of ownership for data warehousing. Organizations running large-scale analytics—whether for human-driven exploration or AI agent queries—can see substantial savings, especially as query volumes grow. Combined with faster execution, this means more work gets done for less money.
Integrated Data Lake Query Engine
One of the standout features of RG instances is the integrated data lake query engine, enabled by default. This allows you to run SQL analytics across both your Redshift data warehouse tables and Amazon S3 data lake using a single engine. No need to switch between systems or move data—just query everything from one place.
Performance for Open Formats
The integrated engine shines with modern data lake formats. For Apache Iceberg tables, RG instances deliver performance up to 2.4x faster than RA3 instances. For Apache Parquet, the speedup is up to 1.5x. This means you can leverage your existing data lake investments while enjoying warehouse-level query speed.
Simplify Operations with a Single System
By unifying data warehouse and data lake querying, RG instances reduce operational complexity. Administrators manage one system, developers write one set of queries, and users get consistent results—whether the data resides in structured warehouse tables or in Amazon S3. This approach lowers total analytics costs for combined workloads.
Designed for Modern Workloads: AI Agents, BI, and ETL
Today’s analytics landscape is evolving rapidly. Data volumes are exploding, and AI agents are now querying data warehouses at scales that dwarf human usage. These autonomous agents demand low-latency responses and high concurrency, which can spiral costs if not managed efficiently. RG instances are built to handle these new patterns.
The combination of faster performance, lower per-query cost, and integrated data lake access makes RG instances particularly suited for:
- BI dashboards requiring sub-second refresh times
- ETL pipelines that need to process large volumes quickly
- Near-real-time analytics for operational decisions
- Autonomous, goal-seeking AI agents that iterate through thousands of queries
RG vs RA3: Choosing the Right Instance
To help you transition, Amazon provides clear upgrade paths from current RA3 instances to recommended RG equivalents. Here’s a quick comparison:
- RA3 XL Plus → RG XLarge: 4 vCPUs, 32 GB memory. Ideal for small cluster departmental analytics.
- RA3 4XL → RG 4XLarge: vCPUs increase from 12 to 16 (1.33:1), memory from 96 GB to 128 GB (1.33:1). Suited for standard production workloads with medium data volumes.
These upgrades offer more compute and memory per node, directly translating to better performance for typical workloads. For exact savings tailored to your usage, we recommend using the AWS Pricing Calculator.
Getting Started with RG Instances
Launching or migrating to RG instances is straightforward. You can:
- Create new clusters via the AWS Management Console
- Use the AWS CLI or API for automation
- Migrate existing RA3 clusters to RG with minimal downtime
The integrated data lake query engine is enabled by default, so you can start querying your data lake immediately. No extra setup required.
With RG instances, Amazon Redshift continues its tradition of innovation—delivering faster, cheaper, and more integrated analytics for the era of AI-driven data exploration. Whether you’re powering human analysts or autonomous agents, RG instances provide the performance and cost efficiency you need to stay ahead.