Contribution to a conference proceedings/Contribution to a book

FZJ-2019-06336

http://join2-wiki.gsi.de/foswiki/pub/Main/Artwork/join2_logo100x88.png The Helmholtz Analytics Toolkit (HeAT) - A Scientific Big Data Library for HPC -

Krajsek, K. (Corresponding author)FZJ* ; Comito, C.FZJ* ; Götz, M. ; Hagemeier, B.FZJ* ; Knechtges, P. ; Siggel, M.

2019
Forschungszentrum Jülich GmbH Zentralbibliothek Jülich
ISBN: 978-3-95806-392-1

Please use a persistent id in citations: http://hdl.handle.net/2128/23601

Abstract: We present HeAT, a scientific big data librarysupporting transparent computation on HPC systems. HeATbuilds on top of PyTorch, which already provides many requiredfeatures like automatic differentiation, CPU and GPU support,linear algebra operations and basic MPI functionality as well asan imperative programming paradigm allowing fast prototypingessential in scientific research. These features are generalized toa distributed tensor with a NumPy-like interface allowing to port existing NumPy algorithms to HPC systems nearly effortlessly.

---

Contributing Institute(s):

1. Jülich Supercomputing Center (JSC)

Research Program(s):

1. 511 - Computational Science and Mathematical Methods (POF3-511) (POF3-511)
2. 512 - Data-Intensive Science and Federated Computing (POF3-512) (POF3-512)
3. HAF - Helmholtz Analytics Framework (ZT-I-0003) (ZT-I-0003)
4. SLNS - SimLab Neuroscience (Helmholtz-SLNS) (Helmholtz-SLNS)

Appears in the scientific report 2019

---

Database coverage:
;

---