cig_logo.png

2018 Rayleigh Tutorial

15 September 2018

Instructor(s)

Nick Featherstone, University of Colorado
Ryan Overdahl, University of Colorado

Description

This workshop will provide an introduction to Rayleigh, an efficiently parallel, pseudo-spectral convection code for spherical and cartesian geometries. An overview of the underlying physics, the parallelization, and performance expectations on different architectures will be provided. Following the overview, participants will gain hands-on experience in running the code and visualizing output.

Technical Requirements

To participate fully in the tutorial, participants should have a basic knowledge of scientific Python and bring with them a reasonably modern laptop with Python 3.x and Matplotlib installed (4GB of ram preferred).

Over the wireless internet, participants will access RMACC Summit Supercomputer to run tutorial exercises.  No additional software installation is necessary.

Rayleigh is written in primarily in Fortran and, optionally, employs simple C++ libraries for directory creation at runtime. In addition, Rayleigh depends on several commonly used external libraries. If you wish to install and run Rayleigh locally, you will need to install the following on your system prior to the tutorial.

  • Fortran compiler (tested GNU, Intel, and IBM)
  • C++ compiler (Optional; tested GNU and Intel)
  • BLAS
  • LAPack
  • FFTW (version 3.x or higher)
  • MPI (for Fortran)
  • Optional: Intel's MKL library provides interfaces to LAPack, BLAS, and FFTW3. A single MKL installation may be used in place of those three, separately installed libraries.

Agenda 

Morning
8:00     Participant Introductions
            Introduction to Rayleigh
            Building and porting the code
Break
            Controlling code physics

Lunch

Afternoon
1:00     Ensemble Mode
            Generating Output
            Plotting Output
Break
            Custom Diagnostics
4:00     Discussions   

5:00     Tutorial Ends

draft 8/10/2018

Resources

More information and link to resources [webpage]

Previous tutorial. [slides][videos]

 

 

 

Location

CU Engineering Center. Enter through courtyard by large stone orbs (south).

Discovery Learning Center, 1B65

Parking: Lot 436   SEE NOTE BELOW ABOUT PARKING

Campus Map

 

!!!! PARKING INFORMATION FOR SATURDAY 15 SEPTEMBER !!!!

DO NOT park in surface lots. Use lot marked above as 436 (NOT 436/494). Park in garage parking at $30.  

There is a football game on Saturday and you will be towed if you park in lot 436/494.

 

 

 

Sign In