Amazon Web Services, Inc. (AWS), an Amazon.com, Inc. company announced the general availability of Amazon Elastic Compute Cloud (Amazon EC2) Hpc6a instances, a new instance type that is purpose-built for tightly coupled high performance computing (HPC) workloads. Hpc6a instances, powered by 3rd Gen AMD EPYC processors, expand AWS's portfolio of HPC compute options and deliver up to 65% better price performance compared to similar compute-optimized Amazon EC2 instances that customers use for HPC workloads today. Hpc6a instances make it even more cost-efficient for customers to scale HPC clusters on AWS to run their most compute-intensive workloads like genomics, computational fluid dynamics, weather forecasting, molecular dynamics, computational chemistry, financial risk modeling, computer-aided engineering, and seismic imaging.

Hpc6a instances are available on demand via a low-cost, pay-as-you-go usage model with no upfront commitments. Organizations across numerous sectors rely on HPC to solve their most complex academic, scientific, and business problems. However, effectively using HPC is expensive because it requires the ability to process large amounts of data, which demands an abundance of compute power, fast memory and storage, and low-latency networking within HPC clusters. Some organizations build infrastructure on premises to run HPC workloads, but that involves expensive upfront capital investment, lengthy procurement cycles, ongoing management of overhead to monitor hardware and keep software up to date, and limited flexibility when the infrastructure inevitably becomes obsolete and must be upgraded.

Customers across many industries run their HPC workloads in the cloud to take advantage of the superior security, scalability, and elasticity it offers. Engineers, researchers, and scientists rely on AWS to run their most complex HPC workloads and choose Amazon EC2 instances with senhanced networking (e.g. C5n, R5n, M5n, and C6gn) to scale tightly coupled HPC workloads that require high levels of inter-instance communications with thousands of interdependent tasks. While the performance of these instances is sufficient for most HPC use cases, as workloads further scale to solve increasingly difficult problems, customers are looking to maximize price performance as they run HPC workloads that can grow to tens of thousands of servers on AWS.

New Hpc6a instances are purpose-built to offer the best price performance for running HPC workloads at scale in the cloud. Hpc6a instances deliver up to 65% better price performance for HPC workloads to carry out complex calculations across a range of cluster sizes—up to tens of thousands of cores. Hpc6a instances are enabled with Elastic Fabric Adapter (EFA)—a network interface for Amazon EC2 instances—by default.

With EFA networking, customers benefit from low latency, low jitter, and up to 100 Gbps of EFA networking bandwidth to increase operational efficiency and drive faster time-to-results for workloads that rely on inter-instance communications. Hpc6a instances are powered by 3rd Gen AMD EPYC processors that run at frequencies up to 3.6 GHz and provide 384 GB of memory. Using Hpc6a instances, customers can more cost-effectively tackle their biggest and most difficult academic, scientific, and business problems with HPC, and realize the benefits of AWS with superior price performance.