Optimizing Distributed Tensor Contractions Using Node-Aware Processor Grids

Andreas Irmler, Raghavendra Kanakagiri, Sebastian T. Ohlmann, Edgar Solomonik, Andreas Grüneis

Research output: Chapter in Book/Report/Conference proceedingConference contribution


We propose an algorithm that aims at minimizing the inter-node communication volume for distributed and memory-efficient tensor contraction schemes on modern multi-core compute nodes. The key idea is to define processor grids that optimize intra-/inter-node communication volume in the employed contraction algorithms. We present an implementation of the proposed node-aware communication algorithm into the Cyclops Tensor Framework (CTF). We demonstrate that this implementation achieves a significantly improved performance for matrix-matrix-multiplication and tensor-contractions on up to several hundreds modern compute nodes compared to conventional implementations without using node-aware processor grids. Our implementation shows good performance when compared with existing state-of-the-art parallel matrix multiplication libraries (COSMA and ScaLAPACK). In addition to the discussion of the performance for matrix-matrix-multiplication, we also investigate the performance of our node-aware communication algorithm for tensor contractions as they occur in quantum chemical coupled-cluster methods. To this end we employ a modified version of CTF in combination with a coupled-cluster code (Cc4s). Our findings show that the node-aware communication algorithm is also able to improve the performance of coupled-cluster theory calculations for real-world problems running on tens to hundreds of compute nodes.

Original languageEnglish (US)
Title of host publicationEuro-Par 2023
Subtitle of host publicationParallel Processing - 29th International Conference on Parallel and Distributed Computing, Proceedings
EditorsJosé Cano, Marios D. Dikaiakos, George A. Papadopoulos, Miquel Pericàs, Rizos Sakellariou
Number of pages15
ISBN (Print)9783031396977
StatePublished - 2023
Event29th International European Conference on Parallel and Distributed Computing, Euro-Par 2023 - Limassol, Cyprus
Duration: Aug 28 2023Sep 1 2023

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume14100 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349


Conference29th International European Conference on Parallel and Distributed Computing, Euro-Par 2023

ASJC Scopus subject areas

  • Theoretical Computer Science
  • General Computer Science


Dive into the research topics of 'Optimizing Distributed Tensor Contractions Using Node-Aware Processor Grids'. Together they form a unique fingerprint.

Cite this