Cluster Overview

Our compute infrastructure consists of two types of resources:

• Non-RCI Servers: Direct-access machines, ideal for debugging and small jobs.
• RCI Cluster: A high-performance SLURM-managed cluster for large-scale training.

This documentation gives you a practical overview of how to start using both systems. For more details, always refer to the official documentation:

• RCI Wiki
• CMP GPU Overview
• CMP Storage

---

Table of contents

• Examples
• Non-RCI Servers
• RCI Cluster
• RCI Interactive Development