Optimization of processor allocation for domain decomposed Monte Carlo calculations
Journal Article
·
· Parallel Computing
- Sandia National Lab. (SNL-NM), Albuquerque, NM (United States)
- Oak Ridge National Lab. (ORNL), Oak Ridge, TN (United States)
- North Carolina State Univ., Raleigh, NC (United States)
Load imbalance plagues domain decomposed Monte Carlo calculations when sources are not uniform. Parallel efficiency for domain decomposed Monte Carlo transport calculations improves through a nonuniform allocation of processors over subdomains. In this work, we optimize the allocation with runtime diagnostics collected during a calibration step, then complete the full calculation. The diagnostic-based approach is compared to implicit filtering, an optimization algorithm for bound constrained noisy optimization problems. Finally, we consider both forward and hybrid radiation transport calculations to measure performance.
- Research Organization:
- Oak Ridge National Laboratory (ORNL), Oak Ridge, TN (United States)
- Sponsoring Organization:
- USDOE; USDOE Office of Science (SC)
- Grant/Contract Number:
- AC05-00OR22725
- OSTI ID:
- 1558566
- Alternate ID(s):
- OSTI ID: 1682480
- Journal Information:
- Parallel Computing, Journal Name: Parallel Computing Journal Issue: C Vol. 87; ISSN 0167-8191
- Publisher:
- ElsevierCopyright Statement
- Country of Publication:
- United States
- Language:
- English
Similar Records
VARIANCE ESTIMATION IN DOMAIN DECOMPOSED MONTE CARLO EIGENVALUE CALCULATIONS
Obtaining Identical Results on Varying Numbers of Processors In Domain Decomposed particle Monte Carlo Simulations
Scalable Domain Decomposed Monte Carlo Particle Transport
Conference
·
2011
·
OSTI ID:1039606
Obtaining Identical Results on Varying Numbers of Processors In Domain Decomposed particle Monte Carlo Simulations
Conference
·
2005
·
OSTI ID:15015123
Scalable Domain Decomposed Monte Carlo Particle Transport
Thesis/Dissertation
·
2013
·
OSTI ID:1116906