AWS launches Graviton5-powered R9g and R9gd instances for data-intensive workloads
AWS has launched its Graviton5-powered R9g and R9gd memory-optimized instances for data-intensive workloads, now available in multiple regions.
The R9g and R9gd instances are designed for applications including Redis OSS, Memcached and Valkey, as well as big data analytics engines, high-throughput web services and containerized microservices on Kubernetes and Amazon EKS. The Graviton5 chips were first announced at AWS re:Invent in December and have already been used in the general-purpose M9g/M9gd and compute-optimized C9g/C9gd instances.
Abib detailed performance improvements: Graviton5 delivers up to 25% better compute performance, 35% faster web application execution and machine learning inference, and 30% faster database processing. The processors are paired with DDR5 8800 memory modules, supporting 8,800 MT/s, a five-times-larger L3 cache and three-times-higher packet-processing performance than Graviton4, reducing idle time for workloads waiting for data.
The instances come in 11 sizes, from medium to metal-48xl with 192 vCPUs and 1,536 GiB of memory. At the top end, they offer up to 100 Gbps network bandwidth and 72 Gbps Amazon Elastic Block Store bandwidth, more than twice that of the previous generation. They also support an Instance Bandwidth Configuration feature, letting users reallocate up to 25% of bandwidth between EBS and virtual private cloud networking.
The key difference between the two is storage: R9g uses attached EBS volumes, while R9gd supports up to 11.4 TB of local NVMe SSD storage for ultra-low-latency workloads such as temporary caches and distributed analytics. Both leverage the AWS Nitro System with the Nitro Isolation Engine for bare-metal-like performance and security.
The new instances are available starting today in US East (N. Virginia, Ohio), US West (Oregon) and Europe (Frankfurt), with On-Demand, Savings Plans, Spot Instances, Dedicated Instances and Dedicated Host pricing options.