AWS Parallel Computing Service features

AWS Parallel Computing Service (AWS PCS) offers a unified set of APIs to help build and operate clusters supporting a range of high performance computing (HPC) and scientific and engineering modeling workloads. Use AWS PCS to stitch together compute, storage, networking, workstations, login nodes, and GPUs. Extend users’ access with your own identity system. Use a feature-rich management console to interactively manage your cluster or automate cluster operations using the AWS SDK and AWS Command Line Interface (AWS CLI).

AWS PCS offers managed compute as well as login and visualization node groups that automatically provision and scale resources in response to user demand, helping optimize resource use and efficiency. AWS PCS supports most Amazon Elastic Compute Cloud (Amazon EC2) instance types, Amazon EC2 capacity allocation schemes, and Multi-AZ deployments.

AWS PCS offers managed service and software updates to deliver the latest features, security patches, and performance improvements with limited cluster downtime. This helps you streamline cluster operations and keep them up to date without significant impact to user productivity. Additionally, AWS PCS provides detailed telemetry and insights so you can monitor your cluster health and performance and troubleshoot user issues effectively.

AWS PCS supports using your own custom AMIs, allowing you to tailor the operating system and preinstalled software to your specific workload needs. You can also make applications available to users through shared file systems or install codes at runtime using instance lifecycle hooks.

AWS PCS offers end users a familiar experience for running and managing their jobs. Whether they are submitting jobs, monitoring performance, or doing interactive data analysis and visualization, their AWS PCS interface is designed to resemble traditional HPC environments. AWS PCS supports your preferred job scheduler, which allows users to operate with the same codes and job scripts between on-premises environments and on AWS. This helps researchers, scientists, and engineers focus on their core research.