Efficient Parallel Scheduling for Sparse Triangular Solvers

Toni Böhnlein, Pál András Papp, Raphael S. Steiner, Albert Jan N. Yzelman

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

We develop and analyze new scheduling algorithms for solving sparse triangular linear systems (SpTRSV) in parallel. Our approach, which we call barrier list scheduling, produces highly efficient synchronous schedules for the forward- and backward-substitution algorithm. Compared to state-of-the-art baselines HDagg [23] and SpMP [16], we achieve a 3.24× and 1.45× geometric-mean speed-up, respectively. We achieve this by obtaining an up to 11× geometric-mean reduction in the number of synchronization barriers over HDagg, whilst maintaining a balanced workload, and by applying a matrix reordering step for locality. We show that our improvements are consistent across a variety of input matrices and hardware architectures.

Original languageEnglish
Title of host publicationProceedings - 2025 IEEE International Parallel and Distributed Processing Symposium Workshops, IPDPSW 2025
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages1263-1265
Number of pages3
ISBN (Electronic)9798331526436
DOIs
StatePublished - 2025
Externally publishedYes
Event2025 IEEE International Parallel and Distributed Processing Symposium Workshops, IPDPSW 2025 - Milan, Italy
Duration: 3 Jun 20257 Jun 2025

Publication series

NameProceedings - 2025 IEEE International Parallel and Distributed Processing Symposium Workshops, IPDPSW 2025

Conference

Conference2025 IEEE International Parallel and Distributed Processing Symposium Workshops, IPDPSW 2025
Country/TerritoryItaly
CityMilan
Period3/06/257/06/25

Bibliographical note

Publisher Copyright:
© 2025 IEEE.

Keywords

  • forward-substitution algorithm
  • Sparse triangular system solving
  • SpTRSV
  • synchronous parallel schedule

Fingerprint

Dive into the research topics of 'Efficient Parallel Scheduling for Sparse Triangular Solvers'. Together they form a unique fingerprint.

Cite this