Tuning and hybrid parallelization of a genetic-based multi-point statistics simulation code

•This work is part of an effort to accelerate geostatistical simulation codes.•We apply acceleration techniques to a genetic-based MPS simulation code.•The acceleration techniques are code optimization and parallelization.•Performance improvements allow us to accelerate up to 100× the execution of t...

Celý popis

Uložené v:
Podrobná bibliografia
Vydané v:Parallel computing Ročník 40; číslo 5-6; s. 144 - 158
Hlavní autori: Peredo, Oscar, Ortiz, Julián M., Herrero, José R., Samaniego, Cristóbal
Médium: Journal Article Publikácia
Jazyk:English
Vydavateľské údaje: Elsevier B.V 01.05.2014
Elsevier
Predmet:
Arrays
Code optimization
Computation
Computer simulation
Convergence
Distributed memory
Enginyeria biomèdica
Genetic algorithms
Genètica
Geostatistics
Multi-point statistics
Optimization
Parallel computing
Parallel processing
Programming
Simulació, Mètodes de
Simulation methods
Stochastic simulation
Àrees temàtiques de la UPC
Stochastic simulation
Parallel computing
Code optimization
Geostatistics
Multi-point statistics
Genetic algorithms
ISSN:0167-8191, 1872-7336
On-line prístup:Získať plný text
Tagy: Pridať tag
Žiadne tagy, Buďte prvý, kto otaguje tento záznam!
Popis
Shrnutí:•This work is part of an effort to accelerate geostatistical simulation codes.•We apply acceleration techniques to a genetic-based MPS simulation code.•The acceleration techniques are code optimization and parallelization.•Performance improvements allow us to accelerate up to 100× the execution of the code.•Samples of simulated results are shown. One of the main difficulties using multi-point statistical (MPS) simulation based on annealing techniques or genetic algorithms concerns the excessive amount of time and memory that must be spent in order to achieve convergence. In this work we propose code optimizations and parallelization schemes over a genetic-based MPS code with the aim of speeding up the execution time. The code optimizations involve the reduction of cache misses in the array accesses, avoid branching instructions and increase the locality of the accessed data. The hybrid parallelization scheme involves a fine-grain parallelization of loops using a shared-memory programming model (OpenMP) and a coarse-grain distribution of load among several computational nodes using a distributed-memory programming model (MPI). Convergence, execution time and speed-up results are presented using 2D training images of sizes 100×100×1 and 1000×1000×1 on a distributed-shared memory supercomputing facility.
Bibliografia:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 23
ISSN:0167-8191
1872-7336
DOI:10.1016/j.parco.2014.04.005