Translational process: Mathematical software perspective

Jack Dongarra; Mark Gates; Piotr Luszczek; Stanimire Tomov

doi:10.1016/j.jocs.2020.101216

Translational process: Mathematical software perspective

Jack Dongarra, Mark Gates, Piotr Luszczek, Stanimire Tomov

Computer Science and Mathematics Div

Research output: Contribution to journal › Article › peer-review

2 Scopus citations

Abstract

Each successive generation of computer architecture has brought new challenges to achieving high performance mathematical solvers, necessitating development and analysis of new algorithms, which are then embodied in software libraries. These libraries hide architectural details from applications, allowing them to achieve a level of portability across platforms from desktops to world-class high performance computing (HPC) systems. Thus there has been an informal translational computer science process of developing algorithms and distributing them in open source software libraries for adoption by applications and vendors. With the move to exascale, increasing intentionality about this process will benefit the long-term sustainability of the scientific software stack.

Original language	English
Article number	101216
Journal	Journal of Computational Science
Volume	52
DOIs	https://doi.org/10.1016/j.jocs.2020.101216
State	Published - May 2021

Keywords

Communication avoiding algorithms
DATAFLOW scheduling runtimes
Hardware accelerators

Access to Document

10.1016/j.jocs.2020.101216

Cite this

@article{715e28e9eb5447f29b48989239b30864,

title = "Translational process: Mathematical software perspective",

abstract = "Each successive generation of computer architecture has brought new challenges to achieving high performance mathematical solvers, necessitating development and analysis of new algorithms, which are then embodied in software libraries. These libraries hide architectural details from applications, allowing them to achieve a level of portability across platforms from desktops to world-class high performance computing (HPC) systems. Thus there has been an informal translational computer science process of developing algorithms and distributing them in open source software libraries for adoption by applications and vendors. With the move to exascale, increasing intentionality about this process will benefit the long-term sustainability of the scientific software stack.",

keywords = "Communication avoiding algorithms, DATAFLOW scheduling runtimes, Hardware accelerators",

author = "Jack Dongarra and Mark Gates and Piotr Luszczek and Stanimire Tomov",

note = "Publisher Copyright: {\textcopyright} 2020 Elsevier B.V.",

year = "2021",

month = may,

doi = "10.1016/j.jocs.2020.101216",

language = "English",

volume = "52",

journal = "Journal of Computational Science",

issn = "1877-7503",

publisher = "Elsevier",

}

TY - JOUR

T1 - Translational process

T2 - Mathematical software perspective

AU - Dongarra, Jack

AU - Gates, Mark

AU - Luszczek, Piotr

AU - Tomov, Stanimire

PY - 2021/5

Y1 - 2021/5

N2 - Each successive generation of computer architecture has brought new challenges to achieving high performance mathematical solvers, necessitating development and analysis of new algorithms, which are then embodied in software libraries. These libraries hide architectural details from applications, allowing them to achieve a level of portability across platforms from desktops to world-class high performance computing (HPC) systems. Thus there has been an informal translational computer science process of developing algorithms and distributing them in open source software libraries for adoption by applications and vendors. With the move to exascale, increasing intentionality about this process will benefit the long-term sustainability of the scientific software stack.

AB - Each successive generation of computer architecture has brought new challenges to achieving high performance mathematical solvers, necessitating development and analysis of new algorithms, which are then embodied in software libraries. These libraries hide architectural details from applications, allowing them to achieve a level of portability across platforms from desktops to world-class high performance computing (HPC) systems. Thus there has been an informal translational computer science process of developing algorithms and distributing them in open source software libraries for adoption by applications and vendors. With the move to exascale, increasing intentionality about this process will benefit the long-term sustainability of the scientific software stack.

KW - Communication avoiding algorithms

KW - DATAFLOW scheduling runtimes

KW - Hardware accelerators

UR - http://www.scopus.com/inward/record.url?scp=85091861254&partnerID=8YFLogxK

U2 - 10.1016/j.jocs.2020.101216

DO - 10.1016/j.jocs.2020.101216

M3 - Article

AN - SCOPUS:85091861254

SN - 1877-7503

VL - 52

JO - Journal of Computational Science

JF - Journal of Computational Science

M1 - 101216

ER -

Translational process: Mathematical software perspective

Abstract

Keywords

Access to Document

Other files and links

Fingerprint

Cite this