Amazon EC2 R9g and R9gd instances powered by AWS Graviton5 processors are now generally available

AWS just released the R9g and R9gd instances, which are memory-optimized EC2 instances built on the latest AWS Graviton5 processors. These new instances deliver up to 25% better compute performance compared to their R8g predecessors, making them a notable upgrade for workloads that demand both raw processing power and large amounts of RAM. The R9g comes with local NVMe storage options via the R9gd variant, giving you flexibility depending on whether you need persistent local disk performance alongside your memory capacity.

Here’s what makes this technically interesting: Graviton5 processors are AWS’s custom-built ARM-based chips, designed specifically for EC2 workloads. Unlike traditional Intel or AMD x86 processors that AWS licenses, these are built in-house and optimized for cloud-native applications. The 25% performance improvement comes from architectural enhancements—better instruction throughput, improved cache efficiency, and refined memory subsystems. For developers familiar with x86, there’s good news: ARM code compatibility is solid, and most mainstream frameworks (Python, Node.js, Java, Go) compile and run efficiently on Graviton. If you’re running containerized workloads, Docker images for ARM are widely available and mature.

The practical value here centers on three main scenarios. First, in-memory databases like Redis, Memcached, or DynamoDB workloads benefit dramatically from the extra compute headroom and memory capacity—you get faster query processing without scaling horizontally. Second, real-time analytics platforms that stream large datasets into memory (think Apache Spark or Clickhouse) can process data faster while handling more concurrent queries. Third, relational databases like PostgreSQL or MySQL with large working sets see measurable gains in throughput and latency, which directly impacts your application’s responsiveness. The R9gd variant is particularly useful when you need both memory and fast local storage for things like temporary query results or caching layers.

If you’re currently running R8g instances or evaluating memory-optimized options, it’s worth testing R9g in your environment. The upgrade path is straightforward—stop your instance, change the instance type, and restart. Cost-wise, Graviton instances traditionally carry a pricing advantage over comparable x86 options (usually 15-20% less), so the 25% performance gain can translate to meaningful cost savings. Start with a non-production workload to validate compatibility, but most teams find the migration friction minimal.