Chapter

Parallel plasma simulation in high performance Fortran

04/2006; DOI:10.1007/BFb0037147 pp.203-212

ABSTRACT Particle-in-cell (PIC) codes are suited for simulating kinetic effects relevant in determining the transport mechanisms in magnetically confined plasmas. A significant improvement of the simulation performance of such codes can be expected from parallelization, e.g., by distributing the particle population among several parallel processors. Parallelization of a hybrid MHD-girokinetic PIC code has been accomplished within the High Performance Fortran (HPF) framework, and tested on the IBM SP2 parallel system, using a particle decomposition technique. The adopted technique requires a moderate effort in porting the code in parallel form and results in intrinsic load balancing and modest inter-processor communication. The performance tests obtained confirm the hypothesis of high effectiveness of the strategy, if targeted towards moderately parallel architectures. Optimal use of resources is also discussed with reference to a specific physics problem.

0 0
 · 
0 Bookmarks
 · 
20 Views
Data provided are for informational purposes only. Although carefully collected, accuracy cannot be guaranteed. The impact factor represents a rough estimation of the journal's impact factor and does not reflect the actual current impact factor. Publisher conditions are provided by RoMEO. Differing provisions from the publisher's actual policy or licence agreement may be applicable.

Keywords

hybrid MHD-girokinetic PIC code
 
IBM SP2 parallel system
 
intrinsic load balancing
 
moderate effort
 
moderately parallel architectures
 
modest inter-processor communication
 
parallel form
 
parallel processors
 
Parallelization
 
particle decomposition technique
 
particle population
 
Particle-in-cell
 
Performance Fortran
 
performance tests
 
simulating kinetic effects relevant
 
simulation performance
 
specific physics problem
 
transport mechanisms