SPLATT: Efficient and Parallel Sparse Tensor-Matrix Multiplication

Shaden Smith, Niranjay Ravindran, Nicholas D. Sidiropoulos, George Karypis

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

78 Scopus citations

Abstract

Multi-dimensional arrays, or tensors, are increasingly found in fields such as signal processing and recommender systems. Real-world tensors can be enormous in size and often very sparse. There is a need for efficient, high-performance tools capable of processing the massive sparse tensors of today and the future. This paper introduces SPLATT, a C library with shared-memory parallelism for three-mode tensors. SPLATT contains algorithmic improvements over competing state of the art tools for sparse tensor factorization. SPLATT has a fast, parallel method of multiplying a matricide tensor by a Khatri-Rao product, which is a key kernel in tensor factorization methods. SPLATT uses a novel data structure that exploits the sparsity patterns of tensors. This data structure has a small memory footprint similar to competing methods and allows for the computational improvements featured in our work. We also present a method of finding cache-friendly reordering and utilizing them with a novel form of cache tiling. To our knowledge, this is the first work to investigate reordering and cache tiling in this context. SPLATT averages almost 30x speedup compared to our baseline when using 16 threads and reaches over 80x speedup on NELL-2.

Original languageEnglish (US)
Title of host publicationProceedings - 2015 IEEE 29th International Parallel and Distributed Processing Symposium, IPDPS 2015
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages61-70
Number of pages10
ISBN (Electronic)9781479986484
DOIs
StatePublished - Jul 17 2015
Event29th IEEE International Parallel and Distributed Processing Symposium, IPDPS 2015 - Hyderabad, India
Duration: May 25 2015May 29 2015

Publication series

NameProceedings - 2015 IEEE 29th International Parallel and Distributed Processing Symposium, IPDPS 2015

Other

Other29th IEEE International Parallel and Distributed Processing Symposium, IPDPS 2015
CountryIndia
CityHyderabad
Period5/25/155/29/15

Bibliographical note

Funding Information:
This work was supported in part by NSF (IIS-0905220, OCI-1048018, CNS-1162405, IIS-1247632, IIP-1414153, IIS-1447788), Army Research Office (W911NF-14-1-0316), Intel Software and Services Group, and the Digital Technology Center at the University of Minnesota. Access to research and computing facilities was provided by the Digital Technology Center and the Minnesota Supercomputing Institute.

Publisher Copyright:
© 2015 IEEE.

Keywords

  • CANDECOMP
  • CPD
  • PARAFAC
  • Sparse tensors

Fingerprint

Dive into the research topics of 'SPLATT: Efficient and Parallel Sparse Tensor-Matrix Multiplication'. Together they form a unique fingerprint.

Cite this