Implementation of Parallel-Computing Algorithms for Systems with Distributed Memory in the EUKLID/V1 Integrated Computer Code

A method of performing parallel computations based on the MPI standard and implemented in the EVKLID-V1 integrated code for safety validation of metal-cooled reactor installations is presented, and the results of speeding-up the calculation of power maneuvering of a lead-cooled reactor are presented...

Full description

Saved in:
Bibliographic Details
Published in:Atomic energy (New York, N.Y.) Vol. 130; no. 5; pp. 273 - 277
Main Authors: Veprev, D. P., Koltashev, D. A., Khachatryan, K. S., Bereznev, V. P., Mosunova, N. A., Strizhov, V. F.
Format: Journal Article
Language:English
Published: New York Springer US 01.09.2021
Springer
Springer Nature B.V
Subjects:
Algorithms
Analysis
Approximation
Codes
Computation
Data exchange
Decomposition
Distributed memory
Energy industry
Hadrons
Heat
Heavy Ions
Multiprocessing
Nuclear Chemistry
Nuclear Energy
Nuclear Physics
Nuclear power plants
Nuclear safety
Physics
Physics and Astronomy
Reactors
System effectiveness
ISSN:1063-4258, 1573-8205
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:A method of performing parallel computations based on the MPI standard and implemented in the EVKLID-V1 integrated code for safety validation of metal-cooled reactor installations is presented, and the results of speeding-up the calculation of power maneuvering of a lead-cooled reactor are presented. The minimum parallelization efficiency was about 50% with 15-fold speed-up on 28 cores. The results obtained show that integrated code EUCLID/V1 can be used effectively in computing systems with distributed memory.
Bibliography:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:1063-4258
1573-8205
DOI:10.1007/s10512-021-00808-6