Conference Presentation (After Call)

FZJ-2019-00929

http://join2-wiki.gsi.de/foswiki/pub/Main/Artwork/join2_logo100x88.png Programming GPU-Accelerated OpenPOWER Systems with OpenACC

Herten, A. (Corresponding author)FZJ*

2018

Please use a persistent id in citations: http://hdl.handle.net/2128/21447

Abstract: This 2h tutorial interactively teaches how to handle the massive computing performance offered by POWER systems with NVLink- attached GPUs – the technology also powering Sierra and Summit, two of the fastest supercomputers in the US. The POWER architecture is presented and the available software stack highlighted. Interactively, programming of the attached GPUs with OpenACC is explored by using real-world examples; the most important OpenACC directives are taught on the way. The resulting GPU-accelerated program can easily be used on other GPU-equipped machines and architectures, by nature of OpenACC's portable approach.

Note: Attached material are the slides shown during the 2h tutorial. The source code shown can be given by the author on request.

---

Contributing Institute(s):

1. Jülich Supercomputing Center (JSC)

Research Program(s):

1. 513 - Supercomputer Facility (POF3-513) (POF3-513)
2. 511 - Computational Science and Mathematical Methods (POF3-511) (POF3-511)

Appears in the scientific report 2018

---

Database coverage:

---