Amazon Web Services (AWS) has officially announced the general availability of its latest generation of memory-optimized cloud instances, the Amazon EC2 R9g and R9gd series. These new instances are powered by the fifth generation of the company’s custom-designed silicon, the AWS Graviton5 processor, marking a significant milestone in the evolution of cloud computing hardware. Designed to address the growing demands of memory-intensive workloads, these instances promise to deliver a substantial leap in compute performance, energy efficiency, and security, building upon the foundation laid by their predecessors.
The introduction of the R9g series comes as AWS continues to push the boundaries of what its internally developed silicon can achieve. According to the company, these new instances offer up to 25% better compute performance compared to the previous Graviton4-based R8g instances. This performance gain is complemented by faster memory, higher network throughput, and increased Amazon EBS bandwidth. By integrating a larger L3 cache and refining the underlying processor architecture, AWS has positioned the R9g series as a powerhouse for organizations looking to optimize their cloud environments for both speed and cost-efficiency.
Targeted Workloads and Performance Capabilities
The R9g instances are engineered specifically to handle high-memory requirements. They are particularly well-suited for a broad spectrum of modern, data-heavy applications. Databases, in-memory caches such as Valkey, Redis, and MemCached, and real-time big data analytics platforms represent the primary use cases for these instances. Furthermore, they are optimized for Linux-based environments, including containerized and micro-service-based architectures powered by Kubernetes, Docker, Amazon EKS, and Amazon ECS. Because of the broad compatibility of the Arm-based architecture, applications developed in popular programming languages—ranging from C/C++ and Rust to Python, Java, .NET Core, Node.js, Ruby, and PHP—can be deployed with minimal friction.
For those requiring high-performance, low-latency storage directly attached to their compute resources, AWS has introduced the R9gd variants. These instances include local NVMe-based SSD block-level storage, providing the necessary scratch space for distributed real-time analytics and large-scale caching tasks. The availability of these two distinct series allows architects to select the right balance of compute, memory, and local storage to suit their specific operational requirements.
Hardware Innovations and the Nitro System
The core of the R9g and R9gd performance is the Graviton5 processor, which introduces several hardware-level improvements over the previous Graviton4 architecture. Beyond the raw silicon, these instances are deeply integrated with the AWS Nitro System. By offloading virtualization, storage, and networking functions to dedicated hardware, the Nitro System enables near-bare-metal performance while maintaining the robust security isolation required for multi-tenant cloud environments.
A critical feature of the R9g and R9gd instances is the implementation of the Nitro Isolation Engine (NIE). First introduced earlier this year with the C9g and M9g instances, the NIE represents a major shift in how cloud providers approach security. It serves as a purpose-built component that enforces strict isolation between virtual machines, mediating all access to memory, CPU register states, and I/O devices through a tightly controlled set of APIs.
Perhaps most significantly, the Nitro Isolation Engine utilizes formal verification—a rigorous mathematical technique that proves the hardware and software behave exactly as intended, rather than relying solely on traditional testing methodologies. This approach has established the AWS Nitro System as a formally verified cloud hypervisor, setting a new industry benchmark for mathematically proven cloud security. For organizations operating in highly regulated industries or those handling sensitive data, the assurance provided by formal verification offers a new level of confidence in the integrity of their cloud infrastructure.
Optimizing Network and Storage Bandwidth
AWS has also introduced Instance Bandwidth Configuration (IBC) for the new R9g and R9gd instances. This feature provides cloud administrators with greater control over their infrastructure, allowing them to adjust the allocation of bandwidth between Amazon EBS and Amazon VPC networking by up to 25%. In complex environments where network traffic patterns are highly variable, this capability allows for granular performance tuning. For instance, a database cluster that requires high EBS throughput for transaction logging can be optimized differently than a web-tier application that is primarily concerned with VPC network latency. This flexibility ensures that the R9g series can adapt to the unique performance profiles of a wide variety of enterprise applications.
Scalability and Migration Path
The R9g and R9gd instances are available in a comprehensive range of sizes, starting from the compact medium size and scaling up to the robust metal-48xl instance. This flexibility ensures that organizations can right-size their deployments, starting small and scaling horizontally or vertically as application demand grows. The consistent performance metrics across these sizes make capacity planning more predictable, while the support for a wide array of operating systems—including Amazon Linux 2023, Ubuntu 22.04+, RHEL 8.4+, and SUSE Linux Enterprise Server—ensures broad compatibility with existing software stacks.
For users currently running workloads on R8g instances, the transition to R9g is designed to be seamless. AWS emphasizes that for the vast majority of applications, no code changes are required. By simply selecting an equivalent R9g instance size, developers can expect to see immediate performance improvements. The same holds true for containerized workloads, where multi-arch container images built for Arm64 can be deployed on R9g instances without modification.
To assist organizations in their migration efforts, AWS offers a suite of resources, including the AWS Graviton Getting Started Guide, which provides best practices for building and optimizing applications on Graviton silicon. For those managing Java-based applications, the AWS Transform tool can automate the migration process from x86 to Graviton, further reducing the effort required to modernize legacy environments. Additionally, the Graviton Savings Dashboard enables organizations to track the cost-efficiency gains resulting from their transition to the new architecture.
Availability and Pricing
Amazon EC2 R9g and R9gd instances are currently available in several key regions, including US East (N. Virginia and Ohio), US West (Oregon), and Europe (Frankfurt). AWS has made these instances available through various procurement models, including On-Demand, Spot Instances, Dedicated Instances, and Dedicated Hosts, as well as through Savings Plans, allowing customers to align their spending with their budget and usage patterns.
As AWS continues to expand its custom silicon footprint, the R9g and R9gd instances represent a strategic step forward for the cloud provider. By combining the efficiency of the Graviton5 processor with the proven security and performance of the Nitro System, AWS aims to provide its customers with a high-performance, secure, and cost-effective foundation for the next generation of cloud-native applications. Customers interested in exploring these new capabilities are encouraged to launch instances via the Amazon EC2 console or consult the detailed documentation provided on the AWS website. Users can also utilize the AWS MCP Server and various plugins to interact with APIs and troubleshoot deployments, ensuring that they have the necessary tools to leverage the full potential of these new compute resources.
