Speaker
Richard Pausch
Description
See the full Abstract at http://ocs.ciemat.es/EPS2018ABS/pdf/P4.2041.pdf
Open, Any-Platform, Leadership-Scale PIC Simulations for Humans
(No Hooks Attached)
R. Pausch1,2, R. Widera1, M. Garten1,2, A. Debus1, I. Goethel1, A. Matthes1,2, B. Worpitz1,2,
S. Starke1, J. Kelling1,2, S. Kossagk1,2, S. Bastrakov1, T. Kluge1, G. Juckeland1,
U. Schramm1,2, T.E. Cowan1,2, M. Bussmann1, A. Huebl1,2
1
Helmholtz-Zentrum Dresden – Rossendorf, Dresden, Germany
2
Technische Universität Dresden, Dresden, Germany
PIConGPU is a fully open, community-driven, 3D and 2D3V particle-in-cell code for the
age of heterogeneous, many-core driven supercomputing. Running from a single source
C++ code base PIConGPU supports both "legacy" CPU architectures as well as modern and
highly parallel architectures such as OpenPOWER, XeonPHI, and Nvidia GPUs.
Especially the latter enable few-hour
turnaround full 3D simulations for
complex studies such as laser-ion
acceleration. The resulting dramatic
demands in post-processing (PBytes+)
are efficiently addressed with
implemented in-situ data reduction
techniques. Those allow asking e.g. for
a wide range of observables relevant
for experiments - up to 100x during the
time frame of an actual beam time.
This is complemented by modern
Visualization of an LWFA simulation and a list of
methods for photon generation, software used.
transport, and X-ray interaction.
Driving, re-using and publishing performance-portable libraries, PIConGPU aims to
provide documented, installable and re-usable software components for the community,
well suited for open data (openPMD) and open science workflows without restrictions.
Latest developments further include a python-centric, extensive framework for specific
experiments, which provides all of the above in an intuitive, non-expert user interface.