RSE training in algorithms for exascale simulations

百亿亿次模拟算法的 RSE 培训

• 批准号: EP/W035782/1
• 负责人: Jemma Shipton
• 金额: $ 4.47万
• 依托单位: UNIVERSITY OF EXETER
• 依托单位国家: 英国
• 项目类别: Research Grant
• 财政年份: 2022
• 资助国家: 英国
• 起止时间: 2022 至 无数据
• 项目状态: 未结题
• 来源: https://gtr.ukri.org/projects?ref=EP%2FW035782%2F1
• 关键词: RSE; training; algorithms; exascale; simulations

The exascale computing landscape in the UK is at an exciting stage,with funding being allocated to novel architectures, new softwareframeworks and innovative algorithms. Through training RSEs we have anopportunity to embed the progress made in these areas into the core ofacademic research and industrial applications, positioning the UK asan international leader in exascale simulations. To grasp thisopportunity is essential that RSEs are trained in algorithms sothat they can take an active part in research in this area. In orderto make informed and creative design choices when writing andoptimising software, RSEs need to have core knowledge of algorithms sothat they can confidently innovate and avoid the common pitfalls thatacademics and industrial partners are already aware of through theirresearch and experience. If this core knowledge is not passed on toRSEs and shared throughout the RSE community, advances made throughthe ExCALIBUR programme research projects risk failing to achievecrucial impact in academic and industrial applications.As described in section 7.3 of the RSE Knowledge Integration LandscapeReview, it is crucial that RSEs have the potential to be activelyinvolved in research. This is one of the key attractions of the jobfor skilled postgraduate students and is essential for retainingskilled RSEs in the role. The Landscape Review acknowledges thatdesign of new algorithms is a research field in itself and requires`strong domain specific knowledge'. We propose to provide training inthis area, alongside opportunities for knowledge exchange andnetworking between academic researchers, postgraduate students, RSEsand industrial partners.We propose to run two three-day workshops and a Summer School toprovide training in state of the art algorithms and core knowledge ofthe underlying foundational mathematical and numerical analysis onwhich they are based. The materials developed in advance of, andduring, these events will be curated and shared online to either beused as stand alone material for individual training or to form thebasis of future summer schools.

英国的亿万级计算领域正处于一个令人兴奋的阶段，资金被分配给新颖的架构，新的软件框架和创新算法。通过培训RSE，我们有机会将这些领域取得的进展嵌入学术研究和工业应用的核心，使英国成为艾级模拟的国际领导者。为了抓住这个机会是必不可少的，RSE的算法培训，使他们能够在这一领域的研究积极参与。为了在编写和优化软件时做出明智和创造性的设计选择，RSE需要拥有算法的核心知识，以便他们可以自信地创新并避免学术界和工业合作伙伴通过他们的研究和经验已经意识到的常见陷阱。如果这一核心知识没有传递给RSE并在整个RSE社区中共享，那么通过ExCALIBUR计划研究项目取得的进展就有可能在学术和工业应用中产生不必要的影响。正如RSE知识整合景观评论第7.3节所述，RSE有潜力积极参与研究至关重要。这是这项工作对熟练研究生的主要吸引力之一，也是留住熟练RSE的关键。景观审查承认，新算法的设计本身就是一个研究领域，需要“强大的领域具体知识”。我们建议在这一领域提供培训，并为学术研究人员、研究生、研究生教育和工业合作伙伴提供知识交流和联网的机会。我们建议举办两个为期三天的研讨会和一个暑期学校，提供最先进的算法和基础数学和数值分析核心知识的培训。在这些活动之前和期间开发的材料将在网上进行策划和分享，作为单独的个人培训材料或形成未来暑期学校的基础。