Scholarlink

Abstract <p>Using MPI for Python, an efficient parallel algorithm for scanning the potential energy surface during diffusion of atomic hydrogen has been developed, providing launch, parallelization and control of calculations. Calculations are organized as processes interacting through a common set of files, which reduces overhead costs during intensive data exchange between processes, minimizes performance losses and simplifies setting the task in the execution queue. The use of the algorithm will contribute to the creation of new structural materials for the nuclear industry and hydrogen energy.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Using MPI for Python to Queue Execution When Scanning Potential Energy Surfaces

  • 期刊论文

  • 发表日期：

    2026年7月20日

查看全文
  • A. V. Romanov,
  • S. D. Kurgalin,
  • S. V. Borzunov,
  • A. G. Potapov

摘要

Abstract

Using MPI for Python, an efficient parallel algorithm for scanning the potential energy surface during diffusion of atomic hydrogen has been developed, providing launch, parallelization and control of calculations. Calculations are organized as processes interacting through a common set of files, which reduces overhead costs during intensive data exchange between processes, minimizes performance losses and simplifies setting the task in the execution queue. The use of the algorithm will contribute to the creation of new structural materials for the nuclear industry and hydrogen energy.