Charm++ parallel programming system

[Akhil langer <akhilanger AT gmail.com>]

Hi Casey,

We have an AMR implementation on Charm. Right now it is just restricted to Advection simulations but we are working on it to make it a generic framework. Right now I do not have performance numbers for AMR. However, ChaNGa group can share performance numbers for the N-body simulations.

Regards,
Akhil

On Wed, Sep 21, 2011 at 3:12 PM, Casey W. Stark <caseywstark AT gmail.com> wrote:

I am interested in tinkering with the charm++ particles and AMR frameworks (or just charm) for cosmological simulations. I realize there is changa, and I'm sure I could leverage some of the N-body work, but I am interested in Eularian hydro. I don't know if enzo developers (http://code.google.com/p/enzo/) have contacted the list before, but I'm sure some of them are interested as well.

Is anyone familiar with the performance of the particles/AMR frameworks well above 1,000 cores? I'm sure this depends a lot on the implementation and details of the solver used, but I figured someone might have tried this before.

Best,

Casey W. Stark

Astrophysics PhD student, UC Berkeley

_______________________________________________
charm mailing list
charm AT cs.uiuc.edu
http://lists.cs.uiuc.edu/mailman/listinfo/charm