A Simple Parallelization Scheme for Extensive Air Shower Simulations
read the original abstract
A simple method for the parallelization of extensive air shower simulations is described. A shower is simulated at fixed steps in altitude. At each step, daughter particles below a specified energy threshold are siphoned off and tabulated for further simulation. Once the entire shower has been tabulated, the resulting list of particles is concatenated and divided into separate list files where each possesses a similar projected computation time. These lists are then placed on a computation cluster where the simulation can be completed in a piecemeal fashion as computing resources become available. Once the simulation is complete, the outputs are reassembled as a complete air shower simulation. The original simulation program (in this case CORSIKA) is in no way altered for this procedure. Verification is obtained by comparisons of 10^(16.5) eV showers produced with and without parallelization.
This paper has not been read by Pith yet.
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.