arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2310.01586cs.DC
收藏

为百亿亿次计算准备应用的经验

Experiences Readying Applications for Exascale

Paul T. Bauman, Reuben D. Budiardja, Dmytro Bykov, Noel Chalmers, Jacqueline Chen, Nicholas Curtis, Marc Day, Markus Eisenbach, Lucas Esclapez, Alessandro Fanfa… 展开作者

Paul T. Bauman, Reuben D. Budiardja, Dmytro Bykov, Noel Chalmers, Jacqueline Chen, Nicholas Curtis, Marc Day, Markus Eisenbach, Lucas Esclapez, Alessandro Fanfarillo, William Freitag, Nicholas Frontiere, Antigoni Georgiadou, Joseph Glenski, Kalyana Gottiparthi, Marc T. Henry de Frahan, Gustav R. Jansen, Wayne Joubert, Justin G. Lietz, Jakub Kurzak, Nicholas Malaya, Bronson Messer, Damon McDougall, Paul Mullowney, Stephen Nichols, Matthew Norman, Thomas Papatheodore, Jon Rood, Philip C. Roth, Sarat Sreepathi, James White, Noah Wolfe

AI总结:

本文总结了为OLCF的Frontier百亿亿次系统准备科学应用的四年经验,涵盖软件可编程性、调优、可移植性等关键问题,评估了早期访问系统的开发作用,并介绍了最佳实践的传播方式与未来系统应用就绪的建议。

摘要:

百亿亿次计算的到来促使人们评估在全球最大超级计算机上开展应用就绪性开发的现有最佳实践。本工作详述了过去四年间,为使科学应用能在橡树岭领导力计算设施(OLCF)的Frontier系统上运行所做准备工作中的观察结果。本文探讨了软件领域的一系列主题,包括可编程性、调优以及可移植性考量,这些都是将应用从现有系统迁移至未来部署系统的关键。一组代表性工作负载为通用系统与软件测试提供了案例研究。我们评估了在多代硬件上使用早期访问系统进行开发的效果。最后,我们讨论了如何通过用户指南、培训等各类活动确定最佳实践并向社区传播,同时给出了确保应用在未来领导力计算系统上具备就绪性的建议。

英文摘要:

The advent of exascale computing invites an assessment of existing best practices for developing application readiness on the world's largest supercomputers. This work details observations from the last four years in preparing scientific applications to run on the Oak Ridge Leadership Computing Facility's (OLCF) Frontier system. This paper addresses a range of topics in software including programmability, tuning, and portability considerations that are key to moving applications from existing systems to future installations. A set of representative workloads provides case studies for general system and software testing. We evaluate the use of early access systems for development across several generations of hardware. Finally, we discuss how best practices were identified and disseminated to the community through a wide range of activities including user-guides and trainings. We conclude with recommendations for ensuring application readiness on future leadership computing systems.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑