scholarly journals Slicing based code parallelization for minimizing inter-processor communication

Author(s):  
Mahmut Kandemir ◽  
Yuanrui Zhang ◽  
Sai Prasanth Muralidhara ◽  
Ozcan Ozturk ◽  
Sri Hari Krishna Narayanan
Keyword(s):  
2021 ◽  
Vol 7 (1) ◽  
pp. 15-20
Author(s):  
Anastasiya V. Shoshina ◽  
Viktor I. Belousov

Mathematical simulation of fast neutron reactors requires high-precision calculations of protection problems based on unstructured meshes. The paper considers and analyzes a parallel version of the ODETTA code (Belousov et al. 2019) with the use of the MPI (Message Passing Interface) library technology (Knyazeva et al. 2006). The code is designed for numerical simulation of neutronic processes in shielding compositions of fast neutron lead cooled reactor plants in normal operating modes, and can be used to calculate the radiation conditions of using structural components and equipment of nuclear power facilities which are assumed to be the sources of and/or exposed to ionizing radiation during their safety justification. The operation of the generated code is compared against the previous version. The MPI-based development of the ODETTA code’s algorithmic part is described. Peculiarities and specific features of the code parallelization are presented, the code modification is given, and respective algorithms are considered. The structure of the ODETTA code based on the MPI is described in brief. The results of using the ODETTA code’s serial and parallel versions in OS Linux (Kostromin 2012) for NRNU MEPhI’s HPC cluster are provided (Savchenko et al. 2020). A comparative analysis is presented for two code implementation options in terms of speed and accuracy of results when using two different clusters and different numbers of nodes for these. Peculiarities of cluster-based calculations are noted.


2010 ◽  
Vol 3 (2) ◽  
pp. 769-817
Author(s):  
I. Wohltmann ◽  
R. Lehmann ◽  
M. Rex

Abstract. ATLAS is a new global Lagrangian Chemistry and Transport Model (CTM), which includes a stratospheric chemistry scheme with 46 active species, 171 reactions, heterogeneous chemistry on polar stratospheric clouds and a Lagrangian denitrification module. Lagrangian (trajectory-based) models have several important advantages over conventional Eulerian models, including the absence of spurious numerical diffusion, efficient code parallelization and no limitation of the largest time step by the Courant-Friedrichs-Lewy criterion. This work describes and validates the stratospheric chemistry scheme of the model. Stratospheric chemistry is simulated with ATLAS for the Arctic winter 1999/2000, with a focus on polar ozone depletion and denitrification. The simulations are used to validate the chemistry module in comparison with measurements of the SOLVE/THESEO 2000 campaign. A Lagrangian denitrification module, which is based on the simulation of the nucleation, sedimentation and growth of a large number of polar stratospheric cloud particles, is used to model the substantial denitrification that occured in this winter.


Sign in / Sign up

Export Citation Format

Share Document