Welcome to Riviera!

Riviera, Colorado State University’s high-performance computing (HPC) system, is a powerful resource designed to accelerate research and innovation across disciplines. Managed by the Data Science Research Institute, Riviera is open to all CSU researchers and students, providing advanced computational capabilities to tackle complex scientific challenges. Whether working on AI-driven projects or bioinformatics research, users can harness Riviera’s robust infrastructure to push the boundaries of discovery. Equipped with 16 A100 GPUs for AI computations and high-memory CPU nodes optimized for bioinformatics, Riviera empowers researchers to process massive datasets, train sophisticated machine learning models, and perform intensive simulations with efficiency. By fostering collaboration and computational accessibility, Riviera strengthens CSU’s commitment to advancing research across diverse fields, enabling breakthroughs in data science, biology, engineering, and beyond.

Accessing Riviera

• Connecting to Riviera
  • Global Protect VPN
  • Credentials
  • Secure Shell
• Frequently Asked Questions

Installing Software

• Setting Up Anaconda
• Installing Python and Packages
• Installing R and Libraries
• Installing Software from Source in Home Directory

Running Software on Riviera

• Riviera Structure
  • Operating System
  • Workload management
  • File Management
  • Compute Nodes
  • Modules
• Slurm
  • What is Slurm?
  • Creating a Slurm SBATCH script
  • What Computing Resources Are Available?
  • Resource Efficiency and Fair Use
• SBATCH Directives
• Containerization
  • What is a container?
  • Singularity
  • Using Images from Docker Hub
  • Binding files to a Singularity Image
  • Building images with Dockerfiles
• Checkpointing Jobs
  • Why Use Checkpoints
  • Implementing Checkpoints
  • Quick Start Guide
  • SRUN Checkpointing Options
  • SBATCH Checkpointing Options
  • SCONTROL
• SBatch Use Cases
  • Job Arrays
  • MPI
  • OpenMP
  • CUDA
  • PyTorch

File Transfer

• Data Transfer and File Management
  • Using SCP for File Transfers
  • Using rsync for Efficient Synchronization
  • Using SFTP for Interactive Transfers
  • Managing Your Storage
  • Additional Resources
  • Summary

Cluster Etiquette

• Prohibited Actions
  • Running Code on the Login node
  • Long Running Interacting Jobs
  • Connecting with an IDE
  • Sensitive Data
  • Long Term File storage
  • Improper Resource Usage
  • Sharing Credentials
• GPU Usage Guidelines
  • When to Use a GPU
  • GPU Node Types and Partitions
  • Available GPU Resources
  • Requesting a GPU in Slurm
  • Monitoring GPU Usage
  • Best Practices
  • GPU Fair Useage
  • Do and Don’t
  • When Code is not Using GPU Effectively