PENELOPE-MPI
RSICC package C00713IBMSP00
October 2002
PENELOPE-2001 (RSICC C00682MNYCP02; NEA-1525/05) was developed by the Universitat de Barcelona and Institut de Tecniques Energetiques, Universitat Politecnica de Catalunya Barcelona, Spain, and the Universidad Nacional de Cordoba, Argentina. The Indiana University Medical Center's Radiation Oncology Department and University Information Technology Services at Indiana University adapted PENELOPE-2001 to increase efficiency of simulations for medical applications.
PENELOPE-MPI extends PENELOPE-2001 by providing MPI parallel drivers; it also adds the ability to read different types of input, such as voxel data.
Penelope-MPI runs under the AIX operating system on IBM SP workstations and can be run in serial mode on IBM. Fortran 90 and C compilers and MPI are required to build executables for the parallel version. The code was tested at RSICC on an IBM SP3 with mpxlf90 and mpcc compilers. GVIEW2D, GVIEW3D, and GVIEWC executables from Penelope-2001 are included to display geometry on the computer screen. They run on personal computers under Microsoft Windows NT or Windows 2000 and are simple and effective tools for debugging geometry definition files.
The physics of the calculations were not changed, and the original documentation is still valid. Both PENELOPE and PENELOPE-MPI are restricted software packages. These packages must be obtained through Radiation Safety Information Computational Center (RSICC), at: http://www-rsicc.ornl.gov/
The goal of the Indiana University research is to improve use of the Gamma Knife, a high precision instrument used to treat intracranial lesions. The Gamma Knife is at this time the most accurate way to deliver radiation to small targeted areas. Physicists in IU's Department of Radiation Oncology are comparing Gamma Knife dose distributions as they are currently calculated (without inhomogeneity and contour corrections) with distributions that would result from highly accurate computational methods using Monte Carlo codes. The goal is to improve treatment planning. The microdosimetry aspect of the research requires a highly complicated transport algorithm and high performance simulation.
