Charmed HPC¶
Charmed HPC is a versatile high-performance computing platform that facilitates the set up and maintenance of HPC clusters. This is done by autonomizing the deployment, integration, and life-cycle management of essential cluster software that enables users to run modern workloads at scale.
Charmed HPC spins up turnkey clusters on a variety of cloud platforms to support write-once, run-anywhere user workloads. It also provides the necessary integrations for GPUs, high bandwidth networking, and shared storage.
The platform enables organizations to focus on obtaining key insights and making data-driven decisions by providing an HPC platform that solves the complexity of deploying and operating an HPC cluster at scale. It is directly beneficial to operations teams and system administrators looking to take full advantage of their HPC hardware, available storage configurations, and high bandwidth networking while minimizing cluster downtime for routine maintenance.
In this documentation¶
Learn more about Charmed HPC: Getting Started tutorial, Underlying projects
Workload management: Deploy Slurm, Manage Slurm, Clean up Slurm, Grafana Dashboards
Storage and Resources: Deploy shared filesystem, GPUs, GRES, Interconnects
Security and Identity: Deploy identity provider, Hardening guidelines, Cryptography
Performance: High availability, Benchmarks
Get started here: A hands-on introduction to Charmed HPC
Step-by-step guides covering key operations for setup, integration, management, and usage
Discussion and clarification of key topics such as high availability, cryptography, and GPUs
Technical information such as underlying projects and dependencies, monitoring, and performance metrics
Project and community¶
Charmed HPC is an open source project of the Ubuntu High-Performance Computing community. Interested in contributing bug fixes, patches, documentation, or feedback? Want to join the Ubuntu HPC community? You’ve come to the right place!
Here’s some links to help you get started with joining the community: