OpenMP: Enabling Massive Node-Level Parallelism - 17th International Workshop on OpenMP, IWOMP 2021, Bristol, UK, September 14-16, 2021, Proceedings

OpenMP: Enabling Massive Node-Level Parallelism - 17th International Workshop on OpenMP, IWOMP 2021, Bristol, UK, September 14-16, 2021, Proceedings
复制标题

OpenMP:实现大规模节点级并行 - 第 17 届 OpenMP 国际研讨会,IWOMP 2021,英国布里斯托尔,2021 年 9 月 14-16 日,会议记录

DOI:
10.1007/978-3-030-85262-7_8
复制
发表时间:
2021
期刊:
--
影响因子:
--
通讯作者:
Schulz H
Schulz H
中科院分区:
--
文献类型:
--
作者:
Schulz H

文献摘要

相似文献

复杂算法的编排需要高水平的自动化,以有效地使用现代硬件。使用OpenMP 5.0进行基于任务的编程是实现这一目标的最佳选择。我们使用三种不同的架构和运行时,在波动方程求解器(ExaHyPE)的背景下研究OpenMP 5.0的任务。我们描述了当前可用运行时中存在的几个任务调度缺陷,演示了它们如何影响性能,并展示了如何解决它们。最后,我们提出对OpenMP标准的扩展。
The orchestration of complex algorithms demands high levels of automation to use modern hardware efficiently. Task-based programming with OpenMP 5.0 is a prominent candidate to accomplish this goal. We study OpenMP 5.0 ’s tasking in the context of a wave equation solver (ExaHyPE) using three different architectures and runtimes. We describe several task-scheduling flaws present in currently available runtimes, demonstrate how they impact performance and show how to work around them. Finally, we propose extensions to the OpenMP standard.