Amazon Web Services Announces General Availability of EC2 R9g and R9gd Instances Powered by AWS Graviton5 Processors
Amazon Web Services (AWS), a subsidiary of Amazon.com, has officially announced the general availability of the Amazon EC2 R9g and R9gd instance families. Powered by the newly introduced AWS Graviton5 processors—touted by the company as the most energy-efficient silicon ever developed by AWS—these memory-optimized instances are engineered to deliver up to a 25% boost in compute performance compared to their predecessor, the Graviton4-based R8g instances. The launch represents a significant milestone in the evolution of custom cloud hardware, addressing the intensifying corporate demand for high-throughput, energy-conscious computing architectures capable of handling demanding enterprise data footprints.
Main Facts and Technological Architecture
The newly launched R9g and R9gd instances are purpose-built for memory-intensive workloads that require rapid data processing, high throughput, and robust reliability. Among the target use cases are distributed databases, in-memory caches such as Redis, Valkey, and Memcached, real-time big data analytics pipelines, and complex Linux-based containerized microservices orchestrated via Kubernetes, Docker, Amazon EKS, and Amazon ECS. Furthermore, the instances provide native, high-performance support for applications developed in widely used programming languages, including C, C++, Rust, Go, Java, Python, .NET Core, Node.js, Ruby, and PHP.
The foundational upgrade of this release is the AWS Graviton5 processor. While maintaining backward compatibility and requiring zero code modifications for the vast majority of applications migrating from R8g hardware, Graviton5 brings several distinct hardware enhancements. These include faster memory subsystems, higher network and Amazon EBS bandwidth limits, and a significantly expanded L3 cache per virtual CPU (vCPU).
A distinguishing feature of the R9g lineup is the inclusion of Instance Bandwidth Configuration (IBC). This capability allows systems administrators to dynamically adjust the allocation of network bandwidth between Amazon Elastic Block Store (Amazon EBS) and Amazon Virtual Private Cloud (Amazon VPC) networking by up to 25%. This fine-grained control enables organizations to tailor hardware profiles precisely to the input/output demands of heavy database queries or high-frequency caching layers.
While the standard R9g instances rely entirely on network-attached EBS storage, the R9gd variants incorporate high-speed, local NVMe-based SSD block-level storage. These NVMe-backed configurations serve as ideal scratch spaces or temporary caching zones for distributed data grids and open-source database engines requiring ultra-low-latency local I/O.
Background Context and Chronology of AWS Silicon
The trajectory of Amazon’s custom silicon program traces back to the 2015 acquisition of Annapurna Labs, a semiconductor design company that laid the groundwork for the initial Graviton processor family introduced in 2018. Over the subsequent seven years, AWS systematically iterated upon its ARM-based architecture, pivoting from general-purpose computing toward specialized, workload-optimized silicon.
- 2018: Introduction of the first-generation AWS Graviton processor, aimed at simple scale-out workloads.
- 2020: Launch of Graviton2, marking a turning point where ARM-based processors achieved price-performance parity with traditional x86 alternatives for mainstream cloud workloads.
- 2021: Release of Graviton3, featuring enhanced floating-point performance and cryptographic processing speeds.
- 2023: Deployment of Graviton4, setting new benchmarks in core density and energy efficiency.
- 2025: Unveiling of the Graviton5 processor and the subsequent commercial rollout of the C9g, M9g, and now R9g/R9gd memory-optimized instance families.
Throughout this evolution, AWS has steadily expanded the scope of its Nitro System—a collection of dedicated hardware and software components that offload virtualization, storage, and networking tasks from the host CPU. The R9g and R9gd instances operate entirely within the Nitro ecosystem, granting applications near-bare-metal performance while maintaining strict security boundaries.
Security Innovation: The Nitro Isolation Engine and Formal Verification
A cornerstone of the R9g and R9gd release is the integration of the Nitro Isolation Engine (NIE). Initially introduced earlier this year alongside the C9g and M9g compute and general-purpose instances, NIE represents a paradigm shift in cloud hypervisor security.
The Nitro Isolation Engine is a purpose-built architectural component responsible for enforcing strict isolation between distinct virtual machines running on the same physical host. It achieves this by aggressively mediating all access to virtual machine memory, CPU register states, and hardware I/O devices through a heavily minimized and audited set of APIs.
Crucially, AWS has subjected the Nitro Isolation Engine to formal verification—a rigorous mathematical technique used to prove that a system’s underlying code and hardware logic adhere strictly to intended specifications across all possible execution states, rather than merely passing empirical test cases. By mathematically demonstrating the integrity of the hypervisor isolation barriers, AWS has established a new verification standard for cloud security, addressing enterprise concerns regarding multi-tenant resource leakage and virtualization vulnerabilities.
Comprehensive Instance Specifications
Both the R9g and R9gd instance families are available across 11 distinct sizing tiers, ranging from a lean entry-level configuration to a massive bare-metal architecture capable of harnessing 192 vCPUs and 1,536 GiB (1.5 TiB) of system memory.
Amazon EC2 R9g Specifications (EBS-Only)
| Instance Size | vCPUs | Memory (GiB) | Instance Storage | Network Bandwidth (Gbps) | EBS Bandwidth (Gbps) |
|---|---|---|---|---|---|
| r9g.medium | 1 | 8 | EBS-Only | Up to 15 | Up to 12 |
| r9g.large | 2 | 16 | EBS-Only | Up to 15 | Up to 12 |
| r9g.xlarge | 4 | 32 | EBS-Only | Up to 15 | Up to 12 |
| r9g.2xlarge | 8 | 64 | EBS-Only | Up to 17 | Up to 12 |
| r9g.4xlarge | 16 | 128 | EBS-Only | Up to 17 | Up to 12 |
| r9g.8xlarge | 32 | 256 | EBS-Only | 17 | 12 |
| r9g.12xlarge | 48 | 384 | EBS-Only | 25 | 18 |
| r9g.16xlarge | 64 | 512 | EBS-Only | 34 | 24 |
| r9g.24xlarge | 96 | 768 | EBS-Only | 50 | 36 |
| r9g.48xlarge | 192 | 1536 | EBS-Only | 100 | 72 |
| r9g.metal-48xl | 192 | 1536 | EBS-Only | 100 | 72 |
Amazon EC2 R9gd Specifications (With Local NVMe SSD Storage)
| Instance Size | vCPUs | Memory (GiB) | Instance Storage (NVMe SSD) | Network Bandwidth (Gbps) | EBS Bandwidth (Gbps) |
|---|---|---|---|---|---|
| r9gd.medium | 1 | 8 | 1 x 59 GB | Up to 15 | Up to 12 |
| r9gd.large | 2 | 16 | 1 x 118 GB | Up to 15 | Up to 12 |
| r9gd.xlarge | 4 | 32 | 1 x 237 GB | Up to 15 | Up to 12 |
| r9gd.2xlarge | 8 | 64 | 1 x 474 GB | Up to 17 | Up to 12 |
| r9gd.4xlarge | 16 | 128 | 1 x 950 GB | Up to 17 | Up to 12 |
| r9gd.8xlarge | 32 | 256 | 1 x 1900 GB | 17 | 12 |
| r9gd.12xlarge | 48 | 384 | 3 x 950 GB | 25 | 18 |
| r9gd.16xlarge | 64 | 512 | 1 x 3800 GB | 34 | 24 |
| r9gd.24xlarge | 96 | 768 | 3 x 1900 GB | 50 | 36 |
| r9gd.48xlarge | 192 | 1536 | 3 x 3800 GB | 100 | 72 |
| r9gd.metal-48xl | 192 | 1536 | 3 x 3800 GB | 100 | 72 |
Industry Implications and Market Analysis
The release of the R9g and R9gd instances arrives at a critical juncture for enterprise cloud consumers. As inflationary pressures and corporate sustainability mandates converge, enterprise Chief Information Officers (CIOs) are under mounting pressure to optimize their total cost of ownership (TCO) while shrinking their carbon footprints.
Industry analysts note that custom ARM-based processors developed directly by hyperscale cloud providers are fundamentally altering the competitive landscape of the semiconductor and cloud computing markets. By decoupling performance scaling from traditional merchant silicon roadmaps, AWS is able to pass substantial cost savings and efficiency gains directly to its customer base.
The 25% performance uplift over Graviton4, combined with enhanced power efficiency, directly impacts data center economics. For large-scale web services, financial institutions managing high-volume transactional databases, and SaaS providers running extensive multi-tenant caching layers, migrating workloads to Graviton5 hardware can yield substantial operational savings. Furthermore, because multi-architecture container images built for the ARM64 instruction set run natively without code alterations, containerized deployments on Amazon EKS and ECS can adopt the new instances with minimal engineering friction.
Migration Pathways and Ecosystem Support
To facilitate rapid adoption, AWS has curated a robust suite of migration tools and resources. Organizations transitioning from older x86 infrastructure or previous-generation Graviton instances can leverage automated migration assistance. For instance, AWS Transform provides automated code transformation capabilities designed to assist developers in migrating Java applications smoothly from x86 architectures to Graviton processors.
For container-based environments, standard Kubernetes deployments, Amazon Elastic Kubernetes Service (EKS), and Amazon Elastic Container Service (ECS) fully support the new hardware profiles, provided that container images are compiled for ARM64 architecture.
Operating system support at launch is extensive, encompassing major commercial and enterprise Linux distributions such as Amazon Linux 2023, Amazon Linux 2, Ubuntu 22.04 and later, Red Hat Enterprise Linux (RHEL) 8.4 and later, SUSE Linux Enterprise Server 15 SP3 and later, and Debian 12 and later.
Availability and Pricing Structure
At launch, Amazon EC2 R9g and R9gd instances are immediately accessible in select global AWS Regions, including:
- US East (N. Virginia)
- US East (Ohio)
- US West (Oregon)
- Europe (Frankfurt)
Customers can procure the new instances through multiple purchasing models designed to accommodate varying budgetary and operational strategies, including Savings Plans, On-Demand instances, Spot Instances, Dedicated Instances, and Dedicated Hosts. Detailed hourly pricing matrices are accessible directly via the official Amazon EC2 pricing portal.
Engineering teams looking to evaluate the hardware can provision instances directly through the Amazon EC2 management console using any compatible ARM-based Amazon Machine Image (AMI). Supplementary technical documentation, the AWS Graviton Getting Started Guide, and cost optimization tools such as the Graviton Savings Dashboard are available to assist developers throughout the evaluation and deployment lifecycle.
Related posts:
