Comparison of High-performance Computing Approaches in the Python Environment for a Five-point Stencil Test Problem
Last edited: 2024-02-18
Repository: http://efurlanm.github.io/bs21/
Stephan Stephany. ORCID 0000-0002-6302-4259
National Institute for Space Research (INPE)
São José dos Campos, SP, Brazil
Abstract. Several of the most important high-performance computing approaches available in the Python programming environment of the LNCC Santos Dumont supercomputer, are compared using a specific test problem. Python includes specific libraries, implementations, development tools, documentation, optimization and parallelization resources. It provides a straightforward way to program using a high level of abstraction, but the parallelization features for exploring multiple cores, processors, or accelerators such as GPUs, are diverse and may not be easily chosen by the user. Serial and parallel implementations of a test problem in Fortran 90 are taken as benchmarks to compare performance. This work is a primer for the use of HPC resources in Python.