Published in Graphics

AMD part of Amazon’s High-Performance Computing

by on12 January 2022

EC2 Hpc6a Instances for HPC Workloads

AMD’s new EPYC chips are behind Amazon’s EC2 Hpc6a Instance Optimised High-Performance Computing (HPC) product.

The Third-Gen AMD EPYC Milan processors deliver up to 65 per cent better price-performance than similar compute-focused Amazon EC2 instances, AMD claims.

Amazon is billing its Hpc6a product as it can help customers run their most compute-intensive HPC workloads like genomics, computational fluid dynamics, weather forecasting, financial risk modelling, EDA for semiconductor design, computer-aided engineering, and seismic imaging.

The tech allows scientists and engineers to solve complex, compute-intensive problems such as computational fluid dynamics (CFD), weather forecasting, and genomics. HPC applications typically require instances with high memory bandwidth, a low latency, high bandwidth network interconnect and access to a fast parallel file system.

Customers across various industries use compute-optimised EFA-enabled Amazon EC2 instances (for example, C5n, R5n, M5n, and M5zn) to maximise a variety of HPC performance workloads. Still, as these workloads scale to tens of thousands of cores, cost-efficiency becomes increasingly essential. We have found that customers are not only looking to optimise performance for their HPC workloads but want to optimise costs as well.

AMD is a preferred partner in the growing HPC industry, with AMD EPYC processors powering 73 supercomputers on the latest Top500 list and holding 70 HPC world records.

To enable predictable thread performance and efficient scheduling for HPC workloads, simultaneous multi-threading is disabled. Thanks to AWS Nitro System, no cores are held back for the hypervisor, making all cores available to your code.

Hpc6a instances introduce targeted features to deliver cost and performance optimisations for customers running tightly coupled HPC workloads that rely on high inter-instance communications. These instances enable EFA networking bandwidth of 100 Gbps and are designed to efficiently scale large tightly coupled clusters within a single Availability Zone.

Amazon said that Hpc6a scales up to 400 nodes (approximately 40,000 cores), with the help of EFA networking, it can maintain roughly 100 per cent scaling efficiency.

Hpc6a instance shows 70 per cent lower cost than c5n, meaning companies can deliver new designs faster and at a lower price when using Hpc6a instances. This means companies can provide unique designs quicker and lower costs when using Hpc6a instances.


Last modified on 12 January 2022
Rate this item
(2 votes)

Read more about: