Summary of AWS’s New R8id, M8id, and C8id Instances
This text details the launch of AWS’s next-generation instances – C8id, M8id, and R8id – built on AMD EPYC 9004 processors and physically connected to the host server. Thes instances represent a significant performance upgrade over the previous (sixth) generation, offering:
* Increased Resources: A threefold increase in vCPUs, memory, and local storage.
* Performance Gains: Up to 43% higher compute performance and a 3.3x increase in memory bandwidth.
* Larger Scale: Instances now scale up to 96xlarge (previously 32xlarge),providing up to 384 vCPUs,3TiB of memory,and 22.8TB of local storage.
* Bare Metal Options: Two bare metal sizes (metal-48xl and metal-96xl) are available for workloads needing direct hardware access.
Here’s a breakdown of each instance type and their optimized workloads:
* C8id: Optimized for compute-intensive tasks like video encoding, image manipulation, and media processing. Benefits from fast, low-latency local storage.
* M8id: Balances compute and memory,suitable for data logging,media processing,and medium-sized data stores needing high-speed local block storage.
* R8id: Designed for memory-intensive applications such as large-scale databases (SQL & NoSQL), in-memory databases, data analytics, and AI inference.
Key Specifications (across 11 sizes per family + 2 bare metal):
The table provides detailed specs for instance sizes ranging from ‘large’ to ’24xlarge’ (the provided excerpt is incomplete). Key metrics include:
* vCPUs: Number of virtual CPUs.
* Memory (GiB): Memory capacity, wiht diffrent allocations for C, M, and R instance types (indicated by *).
* Local NVMe storage (GB): Local solid-state drive storage.
* Network bandwidth (Gbps): Network data transfer rate.
* EBS bandwidth (Gbps): elastic Block Storage data transfer rate.
In essence, AWS is offering a new generation of instances designed to handle increasingly demanding workloads with improved performance, scalability, and adaptability.
Related reading
