GiMMiK - Generating Bespoke Matrix Multiplication Kernels for Accelerators: Application to High-Order Computational Fluid Dynamics

File Description SizeFormat 
Wozniak_GiMMiK.pdfAccepted version674.34 kBAdobe PDFDownload
1-s2.0-S0010465515004506-main.pdfPublished version1.56 MBAdobe PDFDownload
Title: GiMMiK - Generating Bespoke Matrix Multiplication Kernels for Accelerators: Application to High-Order Computational Fluid Dynamics
Author(s): Wozniak, BD
Witherden, FD
Russell, FP
Vincent, PE
Kelly, PHJ
Item Type: Journal Article
Publication Date: 18-Jan-2016
Date of Acceptance: 21-Dec-2015
URI: http://hdl.handle.net/10044/1/28821
DOI: https://dx.doi.org/10.1016/j.cpc.2015.12.012
ISSN: 0010-4655
Publisher: Elsevier
Start Page: 12
End Page: 22
Journal / Book Title: Computer Physics Communications
Volume: 202
Copyright Statement: © 2016 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
Sponsor/Funder: Engineering & Physical Science Research Council (EPSRC)
Engineering & Physical Science Research Council (EPSRC)
Engineering & Physical Science Research Council (EPSRC)
Engineering & Physical Science Research Council (E
Engineering & Physical Science Research Council (EPSRC)
Funder's Grant Number: EP/I006761/1
EP/I00677X/1
EP/I012036/1
EP/L000407/1
EP/K027379/1
Keywords: Matrix Multiplication
GPU
CUDA
Code Generation
GEMM
GiMMiK
cuBLAS
Nuclear & Particles Physics
01 Mathematical Sciences
02 Physical Sciences
08 Information And Computing Sciences
Publication Status: Published
Appears in Collections:Faculty of Engineering
Computing



Items in Spiral are protected by copyright, with all rights reserved, unless otherwise indicated.

Creative Commons