SDSC HPC Training for Advanced Computing Users. – DEV SITE

This GitHub repo will provide materials to support SDSC training classes and workshops.

These events will introduce new and current users to the high performance computing, data, and visualization resources available at SDSC and provide the programming skills necessary to use SDSC resources effectively and efficiently. Workshops offer experienced users more in-depth instruction, including hands-on assistance with their own codes and collaborative discussions with other users and parallel computing experts.

This GitHub repo provides materials to support SDSC training classes and workshops. To learn more about SDSC HPC training programs and events, see: https://www.sdsc.edu/education_and_training/training.html

Release 2023-06

Contents:

• Tapis Deployer
  • User Documentation
  • Developer Documentation
• PPOD_CA
• camera-traps
  • Architectual Overview
  • Quick Start
  • Application Configuration
  • Plugin Configuration
  • Support for NVIDIA
  • Support for ARM CPUs
• Developer Information
  • Using Flatbuffers
  • Updating the flatbuffers messages
  • Plugin Start and Stop Protocol
  • Building and Running under Docker
  • Building and Running the Rust Code
  • Integration Testing
  • Plugin Development
  • Release Procedures
• Acknowledgements
• Event Engine
  • Usage
  • Example
  • Design and Architecture
  • Developing event-engine
• Acknowledgement
• HARP - HPC Application Runtime Predictor
  • Overview
  • WAYS to configure HARP to setup applications for profiling:
  • WAY 1. Configure HARP with TAPIS to configure HARP and applications to profile as container images.
  • WAY 2. Installation-based HARP setup
  • Using HARP to profile an application and predict the execution time
  • Releases
  • Citing HARP
  • Reporting Bugs and Contribution
  • License
• Acknowledgements
• CI Components Catalog
  • Introduction
  • Catalog Schema
  • Deploying the Catalog Locally
  • Application URLs
  • Deployment to Pods
• Acknowledgements
• Constrained Language Models Yield Few-Shot Semantic Parsers
  • Initial set-up
  • Running a command
  • Contributing
  • Trademarks
• Notebooks-101-DEV Tutorial
  • Interactive Video Link:
  • GitHub Repo to training material:
  • License

Expanse Tutorial

Expanse Tutorial:

• Expanse Overview
  • HPC for the long tail of science:
  • System Summary
  • Expanse Scaleable Compute Unit
  • Expanse Connectivity Fabric
  • AMD EPYC 7742 Processor Architecture
• Getting Started on Expanse
  • Expanse Accounts
  • Logging Onto Expanse
  • Obtaining Tutorial Example Code
  • Expanse User Portal
• Expanse 101 (Refactored)
  • Introduction to Running Jobs on the Expanse Supercomputer
  • For more training event and information, see:
• Misc Notes/Updates:
• Managing Accounts on Expanse
  • Expanse-Client Script
  • Using Accounts in Batch Jobs
  • Managing Users on an Account
• Job Charging
• Compiling & Linking Code
  • Supported Compilers
  • AMD Optimizing C/C++ Compiler (AOCC)
  • Using the AOCC Compilers
  • Intel Compilers
  • GNU Compilers
  • PGI Compilers
• Compiling and Running CPU Jobs
  • General Steps: Compiling/Running Jobs
  • Checking Your Environment
  • Hello World (MPI)
  • Hello World (OpenMP)
  • Compiling and Running Hybrid (MPI + OpenMP) Jobs
• Compiling and Running GPU Jobs
  • Using Expanse GPU Nodes
  • Vector Addition (GPU/CUDA)
  • Laplace2D (GPU/OpenACC)
• 7. Running Jobs on Expanse
  • Parallel Models
  • Methods for Running Jobs on Expanse
  • Slurm Resource Manager
• Data and Storage, Globus Endpoints, Data Movers, Mount Points
  • Home File System
  • Parallel Lustre Filesystems.
  • Local Scratch
  • Globus Endpoints