Main Content

You are here


Scientific Software Development: A short course for engineers and reluctant coders

Scientific Software Development: A short course for engineers and reluctant coders

The objective of this course is to teach the components and methods of modern scientific software development by exposure to lectures and hands-on tutorials exemplifying component principles. The workshop will focus on:

  1. the basis and elements of modern scientific software and the development process;
  2. the integration of a modern numerical library, PETSc;
  3. the development of simple multiphysics codes using the MOOSE Framework (built on top of PETSc).

The target audience for the course is graduate students in conventional engineering (MechSE, MatSE, CEE, NPRE, etc.) with a need to write and manage research codes, for instance at a national laboratory or in academia. Others interested in these topics are welcome.

Prerequisites: parallel programming and high-performance computing concepts (CS 420 or equivalent); numerical analysis and applied linear algebra (CS 450, TAM 470, or equivalent). Knowledge of C++ is a necessity, but you can run through this tutorial as an introduction or a refresher beforehand to get up to speed on the language.

Tags: 
Date: 
Tuesday, August 18, 2015