dplasma

DPLASMA is a highly optimized, accelerator-aware, implementation of a dense linear algebra package for distributed heterogeneous systems. It is designed to deliver sustained performance for distributed systems where each node featuring multiple sockets of multicore processors, and if available, accelerators, using the PaRSEC runtime as a backend.

https://github.com/icldisco/dplasma

Science Score: 57.0%

This score indicates how likely this project is to be science-related based on various indicators:

  • CITATION.cff file
    Found CITATION.cff file
  • codemeta.json file
    Found codemeta.json file
  • .zenodo.json file
    Found .zenodo.json file
  • DOI references
    Found 2 DOI reference(s) in README
  • Academic publication links
  • Academic email domains
  • Institutional organization owner
  • JOSS paper metadata
  • Scientific vocabulary similarity
    Low similarity (13.6%) to scientific vocabulary

Keywords

dataflow-programming gpu-acceleration gpu-computing high-performance-computing linear-algebra-library
Last synced: 4 months ago · JSON representation ·

Repository

DPLASMA is a highly optimized, accelerator-aware, implementation of a dense linear algebra package for distributed heterogeneous systems. It is designed to deliver sustained performance for distributed systems where each node featuring multiple sockets of multicore processors, and if available, accelerators, using the PaRSEC runtime as a backend.

Basic Info
  • Host: GitHub
  • Owner: ICLDisco
  • License: other
  • Language: C
  • Default Branch: master
  • Homepage:
  • Size: 5.94 MB
Statistics
  • Stars: 15
  • Watchers: 4
  • Forks: 10
  • Open Issues: 39
  • Releases: 0
Topics
dataflow-programming gpu-acceleration gpu-computing high-performance-computing linear-algebra-library
Created about 4 years ago · Last pushed 8 months ago
Metadata Files
Readme Changelog Contributing License Citation Codeowners Authors

README.rst

DPLASMA_ is the leading implementation of a dense linear algebra package
for distributed, accelerated, heterogeneous systems. It is designed to 
deliver sustained performance for distributed systems where each node featuring
multiple sockets of multicore processors, and if available, accelerators like
GPUs or Intel Xeon Phi. DPLASMA achieves this objective through the state of
the art PaRSEC_ runtime, porting the Parallel Linear Algebra Software for
Multicore Architectures (PLASMA) algorithms to the distributed memory realm.

.. _DPLASMA: https://github.com/icldisco/dplasma
.. _PaRSEC: https://github.com/icldisco/parsec

.. contents:: Table of Contents


Features
========

DPLASMA’s feature set includes:

* Linear systems of equations (Cholesky, LU [inc. pivoting, PP], LDL [prototype]).
* Least squares (QR & LQ).
* Symmetric eigenvalue problem (Reduction to Band [prototype]).
* Level 3 Tile BLAS (GEMM, TRSM, TRMM, HEMM/SYMM, HERK/SYRK, HER2K/SYR2k.
* Covers double real, double complex, single real, and single complex (D, Z, S, C) precisions.
* Provides ScaLAPACK-compatible interface for matrices in F77 column-major layout
* Supports: Linux, Windows, Mac OS X, UN*X (depends on MPI, hwloc)


History
=======

DPLASMA started as the distributed version of ICL PLASMA linear algebra package
implemented on top of the PaRSEC_ runtime, a generic framework for architecture
aware scheduling and management of micro-tasks on distributed many-core heterogeneous
architectures. For a long period, the DPLASMA library was packaged as part of the PaRSEC
project. DPLASMA is now an independent project and uses PaRSEC as a dependency.


Installation
============

Please read the INSTALL.rst_ file for the software dependencies and the installation
instructions. In addition, the `PaRSEC wiki`_ might contain additional information.

.. _INSTALL.rst: https://github.com/ICLDisco/dplasma/blob/master/INSTALL.rst
.. _`PaRSEC wiki`: https://github.com/icldisco/parsec/wiki


Links
=====

When referencing DPLASMA in a publication please use the following reference:
`[10.1109/IPDPS.2011.299]`_: Flexible Development of Dense Linear Algebra Algorithms on Massively Parallel Architectures with DPLASMA [bibtex_].

.. _[10.1109/IPDPS.2011.299]: http://www.icl.utk.edu/node/617
.. _bibtex: http://www.icl.utk.edu/publications/export/bibtex/617
 
If you are interested in this project and want to join the users community, our
mailman will be happy to accept you on the project user (moderated) mailing list
at dplasma-users@icl.utk.edu.

Happy hacking,
  The PaRSEC team.

Owner

  • Name: ICLDisco
  • Login: ICLDisco
  • Kind: organization
  • Location: United States of America

Citation (CITATION.cff)

cff-version: 1.2.0
message: "Please cite this software using the metadata from 'preferred-citation'."
authors:
- family-names: "Bosilca"
  given-names: "George"
  orcid: "https://orcid.org/0000-0003-2411-8495"
- family-names: "Bouteiller"
  given-names: "Aurelien"
  orcid: "https://orcid.org/0000-0001-5108-509X"
- family-names: "Herault"
  given-names: "Thomas"
  orcid: "https://orcid.org/0000-0001-6756-6189"
title: "Distributed Parallel Linear Algebra Software for Multiple Architectures"
url: "https://github.com/icldisco/dplasma"
license: BSD-3-Clause-Open-MPI
preferred-citation:
  type: conference-paper
  title: "Flexible Development of Dense Linear Algebra Algorithms on Massively Parallel Architectures with DPLASMA"
  authors:
  - family-names: "Bosilca"
    given-names: "George"
    orcid: "https://orcid.org/0000-0003-2411-8495"
  - family-names: "Bouteiller"
    given-names: "Aurelien"
    orcid: "https://orcid.org/0000-0001-5108-509X"
  - family-names: "Danalis"
    given-names: "Anthony"
  - family-names: "Faverge"
    given-names: "Mathieu"
  - family-names: "Haidar"
    given-names: "Azzam"
  - family-names: "Herault"
    given-names: "Thomas"
    orcid: "https://orcid.org/0000-0001-6756-6189"
  - family-names: "Kurzak" 
    given-names: "Jakub"
  - family-names: "Langou"
    given-names: "Julien"
  - family-names: "Lemarinier"
    given-names: "Pierre"
  - family-names: "Ltaief"
    given-names: "Hatem"
  - family-names: "Luszczek"
    given-names: "Piotr"
  - family-names: "YarKhan"
    given-names: "Asim"
  - family-names: "Dongarra"
    given-names: "Jack"
  doi: "10.1109/IPDPS.2011.299"
  collection-title: "2011 IEEE International Symposium on Parallel and Distributed Processing Workshops and Phd Forum (IPDPSW)"
  month: 5
  year: 2011
  start: 1432 # First page number
  end: 1441 # Last page number
  issn: "1530-2075"
  editors:
  - name: "IEEE"





GitHub Events

Total
  • Issues event: 10
  • Watch event: 1
  • Issue comment event: 24
  • Push event: 12
  • Pull request event: 30
  • Pull request review event: 32
  • Pull request review comment event: 11
  • Fork event: 1
Last Year
  • Issues event: 10
  • Watch event: 1
  • Issue comment event: 24
  • Push event: 12
  • Pull request event: 30
  • Pull request review event: 32
  • Pull request review comment event: 11
  • Fork event: 1

Packages

  • Total packages: 1
  • Total downloads: unknown
  • Total dependent packages: 0
  • Total dependent repositories: 0
  • Total versions: 0
  • Total maintainers: 4
spack.io: dplasma

DPLASMA is a highly optimized, accelerator-aware, implementation of a dense linear algebra package for distributed heterogeneous systems. It is designed to deliver sustained performance for distributed systems where each node featuring multiple sockets of multicore processors, and if available, accelerators, using the PaRSEC runtime as a backend.

  • Versions: 0
  • Dependent Packages: 0
  • Dependent Repositories: 0
Rankings
Dependent repos count: 0.0%
Average: 27.3%
Dependent packages count: 54.7%
Last synced: 4 months ago

Dependencies

.github/workflows/build_cmake.yml actions
  • actions/checkout v2 composite
  • actions/upload-artifact v3 composite