research data . Dataset . 2020

Scalasca summary analysis of HemeLB_GPU application execution with 129 MPI processes on JUWELS/V100

Brian J. N. Wylie;
Open Access
  • Published: 12 Oct 2020
  • Publisher: Zenodo
Abstract
<p>The CompBioMed HPC CoE flagship application HemeLB (prototype GPU version) was run with a (patched) arteries geometry dataset on JSC&#39;s JUWELS supercomputer, and its execution performance with 129 MPI processes on 32 dual 20-core CPU + quad V100 GPU compute nodes measured by Score-P and analysed by Scalasca (and Vampir)</p>
Subjects
ACM Computing Classification System: ComputingMethodologies_COMPUTERGRAPHICS
free text keywords: Scalasca Score-P CUBE HemeLB JUWELS MPI+GPU
Funded by
EC| CompBioMed
Project
CompBioMed
A Centre of Excellence in Computational Biomedicine
  • Funder: European Commission (EC)
  • Project Code: 675451
  • Funding stream: H2020 | RIA
,
EC| POP2
Project
POP2
Performance Optimisation and Productivity 2
  • Funder: European Commission (EC)
  • Project Code: 824080
  • Funding stream: H2020 | RIA
Download fromView all 3 versions
Zenodo
Dataset . 2020
Provider: Datacite
Zenodo
Dataset . 2020
Provider: Datacite
Any information missing or wrong?Report an Issue