Auszug |
---|
15.06.2023 - > Focus: HPC CFD and GPU usage |
Registration (: to get Zoom link ): upcoming 1st of Maysubscribe to https://listserv.zib.de/mailman/listinfo/training-nhr-bt
Format: online workshop with live demos
Target group: HPC user and CFD developer
Speakers: Dr. Lewin Stein (NHR@ZIB), Dr. Jack Ogaja (NHR@GWDG), Dr. Immo Huismann (www.dlr.de/sp/en)
Program
...
...
Morning session
...
=============================================
10:00 Start & welcome
10:15 05 CFD project planning on a supercomputer (HPC project planning, parallelization)
[Lewin Stein - NHR @ Zuse Institute Berlin]12:10 CFD Parallelization concepts, scaling tests
[Lewin Stein
11:20 Porting OpenFOAM to NHR GPU Systems (Configurations and Opportunities)
[Jack Ogaja - NHR @ Zuse Institute Berlin GWDG]
12:30 Lunch break
Afternoon session
...
=============================================
13:30 CODA developments (example: linear solver running on multiple GPUs) Flow solver CODA - How to simulate aircraft on GPUs
[Immo Huismann - www.dlr.de/sp/en]
??
16:00 14:30 Ansys Fluent - flow through a nozzle (live demo CPU vs. GPU)
[Lewin Stein - NHR @ Zuse Institute Berlin]
15:45 End
Slides
=============================================
will be uploaded after the day01_hpc_cfd_project_planning.pdf
02_parallelization_on_CPU_and_GPU.pdf
03 Porting OpenFOAM to NHR GPU Systems - Jack Ogaja (upcoming soon)
04 Flow solver CODA (confidential)
Live-demo materials/scripts
=============================================
Fluent - 3D subsonic nozzle case: fluent_demo_cpu_vs_gpu.zip
Adapted from: Ansys Fluent Tutorial Guide (Release 2023 R1) - Ch. 8: Modeling Transient Compressible Flow
Older (but freely accessible) 2D version: https://alfaproject.ir/wp-content/uploads/2020/02/part6.pdf