From charlesreid1

Ideas for topics:


  • Introductions
    • Introduction to Unix
    • Introduction to clusters (general)
    • Introduction to CHPC (CHPC clusters, and CHPC people/other stuff)
      • Updraft
      • Ember
      • CHPC people - who does what
      • Research park tour???


  • Software
    • Building your own engineering software
      • Build process: source code to binary
        • Small projects
        • Large projects
        • Huge projects
      • Quality assurance
        • Procedures for software quality assurance
        • Tools for software quality assurance
    • Sharing your engineering software
      • Source code vs. libraries
      • Make systems
    • Libraries
    • 3rd Party Software
      • How to use 3rd party software
        • Linking
        • Including
      • Specific packages for specific functions
        • Combustion packages
        • Linear solver packages
        • Symbolic math packages
        • General-purpose, Matlab-style packages


  • Specific 3rd Party Software Packages
    • General-purpose mathematical environments
      • Matlab
      • Python: Numpy, Scipy, Matplotlib
    • Combustion (Cantera)
    • Linear solver packages
      • Hypre
      • Petsc
      • Trilinos


  • Parallelization
    • MPI and OpenMP
      • Processors vs. threads
    • simple MPI example
    • strategies for parallelization
      • e.g. CFD
      • e.g. linear solvers
      • Karniadakis book


  • Arches
    • overview of relevant combustion theory
    • algorithmic overview
      • tables
      • mixing and reaction, chemistry in general
      • transport equations
      • DQMOM
    • specific details
      • what all that crap is


  • UCF
    • introduction to frameworks in general
    • groups of important functions of the framework
      • e.g. input/output
      • e.g. task parallelization
      • e.g. information passing
      • e.g. solvers
      • go through examples... or at least show people where the examples are!
    • maybe give some more detailed information
      • i.e. finally figure out how exactly an abstract "task" object really works, and how loads are balanced