HPC Recipes for AWS

This repository contains example recipes that demonstrate how to build HPC systems on AWS. The recipes focus on AWS Parallel Computing Service, the managed Slurm service, and also cover AWS ParallelCluster, Research and Engineering Studio, AWS Batch, and several other AWS products.

Getting Started

• If you are new to AWS Parallel Computing Service (AWS PCS), you can watch this short introductory video.
• Discover recipes to help you get started with AWS PCS
• Try launching a HPC cluster in the cloud with just a few clicks.
• Need to set up HPC-ready networking? Choose from a simple example or a more advanced configuration.
• There are also examples of setting up HPC-ready filesystems on AWS for you to try and learn from.
• Prefer other tools? Intro videos are also available for AWS ParallelCluster and Research and Engineering Studio on AWS.

If you just want to explore what all is available, the recipes home page shows you every recipe. Here is an example of what you'll find:

This recipe is named byo_login. It's part of the pcs (Parallel Computing Service) collection. Its tags tell us its a core recipe (i.e. maintained by AWS staff) and that it pertains to EC2, Lambda, and Secrets Manager as well as PCS.

Cross-Region and GovCloud Support

All CloudFormation templates in this repository are designed to work across all AWS partitions, including:

• Standard AWS regions (aws)
• AWS GovCloud regions (aws-us-gov)
• AWS China regions (aws-cn)

This is achieved through:

1. ARN Regionalization: All templates use the AWS::Partition pseudo-parameter for ARNs to ensure they work across all AWS partitions.
2. Console URL Regionalization: All console URLs use the AWS::URLSuffix pseudo-parameter to ensure they generate the correct domain for each partition (e.g., console.amazonaws-us-gov.com for GovCloud).

For more information, see the ARN and Console URL Regionalization Scripts.

Security

See CONTRIBUTING for more information.

AI and Agent Support

This repository includes files designed to help AI systems and coding agents work effectively with the HPC recipes:

• llms.txt - A structured overview of the repository following the llms.txt standard, designed to help Large Language Models understand and navigate the HPC recipes collection.
• AGENTS.md - Development guidelines and context for AI coding agents following the AGENTS.md format, including project structure, best practices, and troubleshooting guidance.

Contributing

We encourage your contributions to this collection. Read up on our contribution process and guidelines. Then, head to Get Started to begin.

License

This repository is licensed under the MIT-0 License. See the LICENSE file.