Scalable Parallel Implementation of Geant4 Using Commodity Hardware and Task Oriented Parallel C

摘要

We describe a scalable parallelization of Geant4 using commodity hardware ina collaborative effort between the College of Computer Science and theDepartment of Physics at Northeastern University. The system consists of aBeowulf cluster of 32 Pentium II processors with 128 MBytes of memory each,connected via ATM and fast Ethernet. The bulk of the parallelization is doneusing TOP-C (Task Oriented Parallel C), software widely used in thecomputational algebra community. TOP-C provides a flexible and powerfulframework for parallel algorithm development, is easy to learn, and isavailable at no cost. Its task oriented nature allows one to parallelize legacycode while hiding the details of interprocess communications. Applicationsinclude fast interactive simulation of computationally intensive processes suchas electromagnetic showers. General results motivate wider applications ofTOP-C to other simulation problems as well as to pattern recognition in highenergy physics.