Meeting Update 03/23/2015 - Eddie

We need an abstract for AstroNum. I thought it would be useful to review our group's previous AstroNum talks, so here is what I found:

2011: Efficient Parallelization for AMR MHD Multiphysics Calculations: Implementation in AstroBEAR

  • AstroBEAR intro
  • Distributed tree, level threading, dynamic load balancing, sweep method, scaling result

2012: Multi-physics with AstroBEAR 2.0

  • AstroBEAR intro
  • Parallelization: threaded AMR and scaling results
  • Multi-physics examples:
    • interaction between magnetized clumps and radiative shocks
    • anisotropic heat conduction
    • magnetic towers
    • disk formation around binaries
    • effect of inhomogeneities within colliding flows

2015: Multi-physics with AstroBEAR 3.0?

  • AstroBEAR intro
  • new code stuff
  • new multi-physics
  • emphasis on my recent jet paper?


2-D Mach Stems

​https://astrobear.pas.rochester.edu/trac/blog/ehansen03202015


Possible Runs for Stampede

  • 3-D pulsed jets could easily use up the rest of the computational time. However, these are not ready to go yet. If I shift my focus, I could probably have this ready to run sometime in April.
  • 3-D Mach stems. Right now, the runs take approximately 3 days each on 120 cores on bluehive which is fine. They are 40 cells/rclump. If we want higher resolution, Stampede could be a good option. We need to think more about exactly which runs we want for the paper, but these would be ready to go right away.
    • An 80 cells/rclump run might use up about 100,000 SUs on Stampede? (rough estimate)

Comments

No comments.