Welcome to the Artemisa User Guide!

Artemisa is a high performance computing infrastructure based on hardware GPGPUs accelerators capable of running advanced scientific tasks. It is supported by advanced network and storage systems.

In these pages it is introduced the hardware, working environment and use recipes for the final users of this infrastructure. Also some step by step tutorials are included, illustrating common use cases.

Contents

• Infrastructure
  • System Overview
• Environment
  • Login to Artemisa
  • User Interface
    • Use UI GPU (gpurun)
  • Worker Nodes use
    • HTCondor
    • Worker Nodes Configuration
    • Submit Description File
    • Job management with HTCondor
  • Storage
  • Utilities
    • Containers
    • CVMFS: HEP Software distribution
    • Local Installed software
• Tutorials
  • 01-User Interface development
    • Virtual environment setup
    • Matrix operation on local GPU
    • Simple classification
    • Summary
  • 02-CPU jobs
    • Local CPU execution
    • Execution in a Worker Node
    • Summary
  • 03-Submit a job to WNs GPU(s)
    • Required files
    • Submit Job
    • Requesting a specific resource
    • Summary
  • 04-Multi-GPU processing
    • Jupyter Notebook preprocessing
    • Multi-GPU job submission
    • Summary
  • 05-Containers
    • Preparation
    • Run container in UI
    • Run container in WN
    • Summary
  • 06-Natural Language Processing
    • Setup
    • Pipelines
• FAQ
• HTCondor Cheat-Sheet
• References
  • Other Materials