July 17th, 2018 by Adam Armstrong
AWS Announces High-Performance EC2 Instances Updates
Today at its annual AWS Summit in New York, AWS announced a new high frequency instance for Amazon Elastic Compute Cloud, Amazon EC2 Z1d. Z1d is designed for users that need the highest single-thread performance with a large amount of memory (such as electronic design automation (EDA), relational databases, and financial simulations). AWS is also releasing two other instances for memory intensive applications, the R5 and R5d.
While the cloud has been adopted more and more there is a lingering thought that, while it is a cost-effective method of storage, the cloud is not generally thought of as fast. This perception isn’t entirely unfounded as early uses of large public cloud, like AWS, were for cheap bulk storage and archive; speed wasn’t needed. But as more applications are born and live, they need performance and AWS is happy to oblige.
With the Z1d instance, AWS leverages custom Intel Xeon Scalable Processors that run up to 4.0 GHz and are powered by sustained all-core Turbo Boost. Z1d is for applications that need very high per-core performance. Z1d will come in six sizes ranging from 2 vCPUs with 16GiB of memory and 75GB of NVMe storage all the way up to 48 vCPUs with 384GiB of memory and 1.8TB of NVMe storage. Bandwidth will range from 10Gb/s to 25Gb/s.
For memory-optimized instances, AWS will be offering the R5 with up to 3.1GHz powered by sustained all-core Turbo Boost and the R5d that is equipped with local NVMe storage (up to 3.6TB). These instances are designed for high-performance databases, distributed in-memory caches, in-memory analytics, and big data analytics. The instances will leverage custom Intel Xeon Platinum 8000 Series CPUs. They R5 instances also come in six sizes, this time ranging from 2 vCPUs to 96 vCPUs, network bandwidth runs from 10Gb/s to 25Gb/s.
AWS hasn’t given a specific time to availability outside of saying “soon.”
Discuss this story