10 Key Benefits of Amazon Redshift's New Graviton-Powered RG Instances

Amazon Redshift has long been the go-to cloud data warehouse, delivering enterprise-grade performance at a fraction of on-premises costs. Now, with the introduction of AWS Graviton-based RG instances, Redshift takes another leap forward. These new instances combine a custom-designed integrated data lake query engine with the efficiency of Graviton processors, offering up to 2.2x faster performance and 30% lower price per vCPU compared to RA3 instances. Whether you're running BI dashboards, ETL pipelines, or supporting AI agents that query at massive scale, RG instances are built to handle the load. Below, we break down the 10 key things you need to know about this exciting new offering.

1. What Are Amazon Redshift RG Instances?

Amazon Redshift RG instances are a new family of compute nodes powered by AWS Graviton processors. They are designed to deliver better price-performance for data warehouse and data lake workloads. Compared to the previous RA3 instances, RG instances offer up to 2.2x faster query execution and a 30% lower price per vCPU. This makes them an attractive option for organizations looking to optimize costs without sacrificing speed. The integrated data lake query engine is enabled by default, allowing you to run SQL analytics across both Redshift managed storage and Amazon S3 data lakes from a single engine.

10 Key Benefits of Amazon Redshift's New Graviton-Powered RG Instances — Source: aws.amazon.com

2. Up to 2.2x Faster Performance Than RA3 Instances

One of the headline benefits of RG instances is their raw performance. Benchmarks show they can run data warehouse workloads up to 2.2 times faster than comparable RA3 instances. This speed improvement comes from the combination of Graviton processors and architectural optimizations. For workloads like BI dashboards, near-real-time analytics, and ETL pipelines, this means lower latency and faster insights. AI agents that query your data warehouse at scale also benefit greatly from the reduced response times, making RG instances ideal for high-frequency query environments.

3. 30% Lower Price per vCPU

Cost efficiency is a major driver behind the new RG instances. Despite delivering superior performance, they come with a 30% reduction in price per vCPU compared to RA3 instances. This is achieved through the use of energy-efficient Graviton chips and optimized memory configurations. For example, the rg.4xlarge instance offers 16 vCPUs and 128 GB of memory, replacing the ra3.4xlarge (12 vCPUs, 96 GB) at a lower cost per vCPU. When combined with the performance gains, the overall price-performance ratio is significantly improved, helping you do more with your analytics budget.

4. Integrated Data Lake Query Engine Boosts Iceberg and Parquet

RG instances come with a built-in data lake query engine that accelerates queries on popular open table formats. For Apache Iceberg, performance is up to 2.4x faster than RA3 instances; for Apache Parquet, it's up to 1.5x faster. This engine runs within the Redshift service, so you don't need separate tools or compute resources to query data lakes. It automatically optimizes data access patterns and uses the same SQL engine as your warehouse tables. This integration means you can analyze diverse datasets in S3 without sacrificing query performance.

5. Seamless Querying Across Warehouse and Data Lake

With the integrated query engine, you can run SQL queries that span both Redshift managed tables and Amazon S3 data lakes from a single endpoint. There's no need to move data or set up federated queries. This simplifies your data architecture and reduces operational overhead. You can use familiar SQL constructs to join data from warehouse tables with data stored in Iceberg, Parquet, or other formats in S3. This capability is especially valuable for organizations that want to combine structured, frequently accessed data with cost-effective storage of diverse datasets.

6. Ideal for AI Agents and High-Volume Analytics

As AI agents become more prevalent, they generate query volumes that dwarf typical human usage. RG instances are well-suited for these workloads because they handle high concurrency and low-latency requirements efficiently. The performance improvements from Graviton and the data lake engine mean agent queries return results faster, reducing compute time and cost. Whether you're running machine learning pipelines, automated reporting, or real-time decision support, RG instances provide the speed and scalability needed to support autonomous, goal-seeking AI systems.

7. Comparison: RG vs RA3 - Choosing the Right Instance

To help you evaluate, here is a quick comparison of the recommended migration paths from RA3 to RG instances:

ra3.xlplus (4 vCPU, 32 GB) → rg.xlarge (4 vCPU, 32 GB) – Suitable for small clusters and departmental analytics.
ra3.4xlarge (12 vCPU, 96 GB) → rg.4xlarge (16 vCPU, 128 GB) – 1.33:1 vCPU and memory increase – standard production workloads with medium data volumes.

For larger instances, similar ratios apply. The price per vCPU is 30% lower across the board. Use the AWS Pricing Calculator with your specific workload patterns to estimate savings.

8. Simplified Operations with Single Query Engine

By combining warehouse and data lake querying into one engine, RG instances simplify your operational landscape. You no longer need to maintain separate query engines or tools for different data sources. This reduces complexity in monitoring, tuning, and security management. The integrated engine is enabled by default on all new RG clusters, so you benefit immediately without additional configuration. This unification also makes it easier to enforce consistent data governance policies across all your data.

9. How to Launch or Migrate to RG Instances

Getting started with RG instances is straightforward. You can launch new clusters directly from the AWS Management Console, AWS CLI, or AWS API. For existing Redshift clusters, you can migrate by changing the instance type. The integrated data lake query engine is automatically enabled, so no manual setup is required. AWS recommends testing your workloads on a small RG cluster first to validate performance and cost savings before migrating production environments. The migration process is designed to minimize downtime and preserve your existing data.

10. Taking Advantage of AWS Graviton Benefits

AWS Graviton processors are custom-built by AWS using 64-bit Arm cores. They deliver better price-performance than x86-based instances for many workloads. By adopting Graviton, RG instances also benefit from AWS's ongoing investment in chip design, which means future improvements will come automatically. The energy efficiency of Graviton also helps reduce your cloud carbon footprint. If your organization is committed to sustainability, RG instances are a solid choice for data analytics workloads.

Amazon Redshift's new RG instances represent a significant step forward in price-performance for cloud data warehousing and data lake analytics. With up to 2.2x faster performance, 30% lower cost per vCPU, and an integrated query engine, they are built to handle the most demanding analytics and AI agent workloads. Whether you are migrating from older RA3 instances or starting fresh, RG instances offer a compelling blend of speed, efficiency, and simplicity. Try them today and see how they can transform your data analytics operations.