Online data analysis and reduction: An important Co-design motif for extreme-scale computers. (November 2021)

• Record Type: Journal Article
• Authors: Foster, Ian; Ainsworth, Mark; Bessac, Julie; Cappello, Franck; Choi, Jong; Di, Sheng; Di, Zichao; Gok, Ali M; Guo, Hanqi; Huck, Kevin A; Kelly, Christopher; Klasky, Scott; Kleese van Dam, Kerstin; Liang, Xin; Mehta, Kshitij; Parashar, Manish; Peterka, Tom; Pouchard, Line; Shu, Tong; Tugluk, Ozan; van Dam, Hubertus; Wan, Lipeng; Wolf, Matthew; Wozniak, Justin M; Xu, Wei; Yakushin, Igor; Yoo, Shinjae; Munson, Todd
• Other Names: Germann Tim guest-editor.
• Is Part Of: International journal of high performance computing applications. Volume 35:Number 6(2021)
• Page Start: 617
• Page End: 635
• Publication Date: 2021-11
• Subjects: Data analysis -- in situ -- exascale computing -- online data analysis and reduction; High performance computing -- Periodicals; Supercomputers -- Periodicals; 004.1105
• DOI: 10.1177/10943420211023549
• Languages: English
• ISSNs: 1094-3420

Abstract

A growing disparity between supercomputer computation speeds and I/O rates means that it is rapidly becoming infeasible to analyze supercomputer application output only after that output has been written to a file system. Instead, data-generating applications must run concurrently with data reduction and/or analysis operations, with which they exchange information via high-speed methods such as interprocess communications. The resulting parallel computing motif, online data analysis and reduction (ODAR), has important implications for both application and HPC systems design. Here we introduce the ODAR motif and its co-design concerns, describe a co-design process for identifying and addressing those concerns, present tools that assist in the co-design process, and present case studies to illustrate the use of the process and tools in practical settings.