Welcome to the documentation for GRACE-HPC!

Green Resource for Assessing Carbon & Energy in HPC

GRACE-HPC is a Python toolkit for estimating the energy consumption and carbon emissions of SLURM-based HPC jobs. Calculate your computational carbon footprint through a CLI, Python functions, and an interactive Jupyter Notebook interface.

• GitHub: https://github.com/Elliot-Ayliffe/GRACE-HPC

• Python Versions: 3.11, 3.12

• Open Source: MIT License

• Issues: https://github.com/Elliot-Ayliffe/GRACE-HPC/issues

• Contact: elliot.ayliffe@gmail.com

Background:

• Introduction
  • Motivation
  • Terminology
• About GRACE-HPC
  • What is GRACE-HPC?
  • How to use it?
  • Features
• Methodology
  • Pull Job Data
  • Energy Consumption
  • Scope 2 Emissions
  • Scope 3 Emissions
  • Contextual Equivalents
  • Assumptions & Limitations
  • Acknowledgements

User Guide:

• Installation
  • Virtual Environments
  • From PyPI
  • From Source
  • Dependencies
  • Double Check
• Configuration
  • Example hpc_config.yaml
• Inputs & Outputs
  • Input Arguments
  • Output Data
• Command-line Interface (CLI)
  • Argument Options
  • Example Commands
  • Example Terminal Output
• Python Usage
  • Import the Function
  • Function Arguments
  • Run the Engine
  • Function Returns
• Jupyter Interface (Interactive)
  • Setting Up the .ipynb Notebook
  • Launch the Interface
  • Notebook Demo
• API Reference
  • gracehpc_run(args)
  • jupyter_UI()

Help:

• FAQs & Development
  • FAQs
  • Future Development