Skip to main navigation Skip to search Skip to main content

Scalable Block-Sparse Matrix Multiplication Using Template Task Graphs

  • University of Tennessee
  • Virginia Polytechnic Institute and State University

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

Abstract

Block-sparse matrix operations are a special case of general sparse algebra where the matrix is sparsely populated with dense blocks, e.g., in sparse tensor algebra for quantum chemistry. One of the challenges of implementing distributed matrix multiplication C=A×B in general is the management of communication flows since both input matrices A and B are readily available and must be distributed to the processes computing the relevant blocks of C. In this paper, we propose an addition to the Template Task Graph programming model that allows applications to constrain the execution of tasks using a flexible API. We show that such constraints can be used in a pure dataflow model to replace artificial control flow with a more structured approach. In the context of sparse matrix multiplication, we found that constraints allow us to limit the number of concurrent communications and thus avoid creating a bottleneck in the network.

Original languageEnglish
Title of host publicationAsynchronous Many-Task Systems and Applications - 3rd International Workshop, WAMTA 2025, Proceedings
EditorsPatrick Diehl, Qinglei Cao, Thomas Herault, George Bosilca
PublisherSpringer Science and Business Media Deutschland GmbH
Pages120-132
Number of pages13
ISBN (Print)9783031971952
DOIs
StatePublished - 2026
Event3rd International Workshop on Asynchronous Many-Task Systems and Applications, WAMTA 2025 - St. Louis, United States
Duration: Feb 19 2025Feb 21 2025

Publication series

NameLecture Notes in Computer Science
Volume15690 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference3rd International Workshop on Asynchronous Many-Task Systems and Applications, WAMTA 2025
Country/TerritoryUnited States
CitySt. Louis
Period02/19/2502/21/25

Keywords

  • Scheduling Constraints
  • Sparse Matrix Multiplication
  • Template Task Graph

Fingerprint

Dive into the research topics of 'Scalable Block-Sparse Matrix Multiplication Using Template Task Graphs'. Together they form a unique fingerprint.

Cite this