GraphScale

• Verfasst von: Dann, Jonas [VerfasserIn]
• Verfasst von: Ritter, Daniel [VerfasserIn]
• Verfasst von: Fröning, Holger [VerfasserIn]
• Titel: GraphScale
• Titelzusatz: scalable Processing on FPGAs for HBM and large graphs
• Verf.angabe: Jonas Dann, Daniel Ritter, Holger Fröning
• E-Jahr: 2024
• Jahr: March 2024
• Umfang: 23 S.
• Illustrationen: Illustrationen
• Fussnoten: Veröffentlicht: 23. März 2024 ; Gesehen am 09.12.2024
• Titel Quelle: Enthalten in: Association for Computing MachineryACM transactions on reconfigurable technology and Systems
• Ort Quelle: New York, NY : ACM Press, 2008
• Jahr Quelle: 2024
• Band/Heft Quelle: 17(2024), 2, Artikel-ID 22, Seite 1-23
• ISSN Quelle: 1936-7406
• ISSN Quelle: 1936-7414
• DOI: doi:10.1145/3616497
• Datenträger: Online-Ressource
• Sprache: eng
• K10plus-PPN: 1911194089

Abstract

Recent advances in graph processing on FPGAs promise to alleviate performance bottlenecks with irregular memory access patterns. Such bottlenecks challenge performance for a growing number of important application areas like machine learning and data analytics. While FPGAs denote a promising solution through flexible memory hierarchies and massive parallelism, we argue that current graph processing accelerators either use the off-chip memory bandwidth inefficiently or do not scale well across memory channels.In this work, we propose GraphScale, a scalable graph processing framework for FPGAs. GraphScale combines multi-channel memory with asynchronous graph processing (i.e., for fast convergence on results) and a compressed graph representation (i.e., for efficient usage of memory bandwidth and reduced memory footprint). GraphScale solves common graph problems like breadth-first search, PageRank, and weakly connected components through modular user-defined functions, a novel two-dimensional partitioning scheme, and a high-performance two-level crossbar design. Additionally, we extend GraphScale to scale to modern high-bandwidth memory (HBM) and reduce partitioning overhead of large graphs with binary packing.