Parallel implicit Monte Carlo in C++
The authors are developing a parallel C++ Implicit Monte Carlo code in the Draco framework. As a background and motivation for the parallelization strategy, they first present three basic parallelization schemes. They use three hypothetical examples, mimicking the memory constraints of the real world, to examine characteristics of the basic schemes. Next, they present a two-step scheme proposed by Lawrence Livermore National Laboratory (LLNL). The two-step parallelization scheme they develop is based upon LLNL`s two-step scheme. The two-step scheme appears to have greater potential compared to the basic schemes and LLNL`s two-step scheme. Lastly, they explain the code design and describe how the functionality of C++ and the Draco framework assist the development of a parallel code.
- Research Organization:
- Los Alamos National Lab. (LANL), Los Alamos, NM (United States)
- Sponsoring Organization:
- USDOE, Washington, DC (United States)
- DOE Contract Number:
- W-7405-ENG-36
- OSTI ID:
- 304119
- Report Number(s):
- LA-UR-98-2135; CONF-981207-; ON: DE99000753; TRN: AHC29904%%195
- Resource Relation:
- Conference: ISCOPE `98: international symposium on computing in object oriented parallel environments, Santa Fe, NM (United States), 8-11 Dec 1998; Other Information: PBD: [1998]
- Country of Publication:
- United States
- Language:
- English
Similar Records
Comparison of four parallel algorithms for domain decomposed implicit Monte Carlo
Comparison of four parallel algorithms for domain decomposed implicit Monte Carlo.