Computing Resources

The Bay-Delta SCHISM model is computationally intensive and requires access to high-performance computing (HPC) resources to run efficiently. In this section, we provide:

A cursory introduction to HPC resources and how to use them for running the Bay-Delta SCHISM model,

• Intro to Clusters and High Performance Computing
  • Anatomy of a Cluster
  • Figure: Cluster Components
  • Compilation
  • Job Submission
  • Compute Nodes and Message Passing
  • Input/Output and Scribes
  • Example Slurm Launch File
  • Example Shell Script to Run SCHISM
  • Intro to Batch Cloud Cluster
  • Figure: Azure Batch Cluster Components

A look at one example of cloud computing with Azure batch:

• Running on Azure Batch
  • Setup Azure
  • Upload Applications
  • File Transfer Commands
  • Ensure Batch Quota
  • References

And some step-by-step guides for running Hello SCHISM and the Bay-Delta SCHISM model on Azure Batch:

• Step-by-step: HelloSCHISM on Azure
  • Initial Setup
  • Uploading Model Input Files
  • Set up Model Run Config
  • Run Simulation!
• Step-by-step: Bay-Delta SCHISM on Azure
  • Initial Setup
  • Uploading Model Input Files
  • Running Model Simulation