Skalierbarkeit von Molekulardynamiksimulationen auf heterogener Hardware
Molekulardynamiksimulationen (MD-Simulationen) haben sich zu einer unverzichtbaren Forschungsmethode in Biochemie und Materialwissenschaft entwickelt. Da die Eingabegröße für MD-Simulationen häufig durch die Größe von Molekülen beschränkt ist, streben MD-Anwendungen strong scaling an. Daraus folgend ist der Rechenaufwand pro Rechenknoten sehr gering, weswegen moderne MD-Anwendungen latenz- und synchronisationskritisch sind. Um das Potential heterogener Hardware auszuschöpfen, müssen sich die Anwendungen generisch an Unterschiede in Latenz, Durchsatz und Speicherhierarchie anpassen, die durch die Kombination von Multi- und Manycore-CPUs mit GPUs und Beschleunigern entstehen.
Veröffentlichungen
- David Haensel, Laura Morgenstern, Andreas Beckmanns, Ivo Kabadshow, Holger Dachsel, "Eventify: Event-Based Task Parallelism for Strong Scaling", in Proceedings of the Platform for Advanced Scientific Computing Conference, 2020 @inproceedings{Morgenstern:2020:PASC,
author = {David Haensel and Laura Morgenstern and Andreas Beckmanns and Ivo Kabadshow and Holger Dachsel},
title = {Eventify: Event-Based Task Parallelism for Strong Scaling},
year = {2020},
isbn = {9781450379939},
publisher = {Association for Computing Machinery},
address = {New York, NY, USA},
url = {https://doi.org/10.1145/3394277.3401858},
doi = {10.1145/3394277.3401858},
booktitle = {Proceedings of the Platform for Advanced Scientific Computing Conference},
articleno = {14},
numpages = {10},
keywords = {scalability},
location = {Geneva, Switzerland},
series = {PASC '20}
- Laura Morgenstern, David Haensel, Andreas Beckmann, Ivo Kabadshow, "NUMA-Awareness as a Plug-In for an Eventify-Based Fast Multipole Method", in Computational Science - ICCS 2020, 428-441, 2020 @InProceedings{Morgenstern:2020:ICCS,
author="Laura Morgenstern and David Haensel and Andreas Beckmann and Ivo Kabadshow",
title="NUMA-Awareness as a Plug-In for an Eventify-Based Fast Multipole Method",
booktitle="Computational Science -- ICCS 2020",
year="2020",
keywords = {scalability},
publisher="Springer International Publishing",
address="Cham",
pages="428--441",
isbn="978-3-030-50436-6"
- Laura Morgenstern, Ivo Kabadshow, "The Rocky Road to Tasking: Task Queues Reloaded", 2020
- Laura Morgenstern, "Tasking Meets GPUs: A Closer Look at Locking", 2020
- Laura Morgenstern, Ivo Kabadshow, Matthias Werner, "GPU-Tasking à la Carte? Eventify Meets GPUs", 2020
@misc{morgenstern:2020:ISC,
author={Laura Morgenstern and Ivo Kabadshow and Matthias Werner},
title={{GPU-Tasking à la Carte? Eventify Meets GPUs}},
url={{https://2020.isc-program.com/presentation/?id=phd113&sess=sess328}},
howpublished={{Talk and Poster at PhD Forum at ISC High Performance 2020}},
keywords= {scalability},
year={2020}
- Bartosz Kohnke, Thomas R. Ullmann, Andreas Beckmann, Ivo Kabadshow,
David Haensel, Laura Morgenstern, Plamen Dobrev, Gerrit Groenhof, Carsten Kutzner, Berk Hess, Holger Dachsel, Helmut Grubmüller, "A scalable and versatile Fast Multipole Method for biomolecular simulation", in Software for Exascale Computing - SPPEXA 2016-2019, Springer Lecture Notes in Computer Science and Engineering, 2019 @inproceedings{Morgenstern:2019:LNCSE,
title={A scalable and versatile Fast Multipole Method for biomolecular simulation},
author={Bartosz Kohnke and Thomas R. Ullmann and Andreas Beckmann and Ivo Kabadshow and
David Haensel and Laura Morgenstern and Plamen Dobrev and Gerrit Groenhof and Carsten Kutzner and Berk Hess and Holger Dachsel and Helmut Grubmüller},
booktitle={Software for Exascale Computing -- SPPEXA 2016-2019, Springer Lecture Notes in Computer Science and Engineering},
year={2019},
keywords = {scalability},
annote = {accepted}
- Laura Morgenstern, Andreas Beckmann, Ivo Kabadshow, "Towards Unified Tasking on CPUs and GPUs", 2019
@misc{morgenstern:2019:GTC,
author={Laura Morgenstern and Andreas Beckmann and Ivo Kabadshow},
title={{Towards Unified Tasking on CPUs and GPUs}},
url={https://www.nvidia.com/en-us/gtc/poster-gallery/hpc-and-supercomputing/#img18},
howpublished={{Poster at GPU Technology Conference 2019, San Jose}},
keywords= {scalability},
year={2019}
- Laura Morgenstern, Andreas Beckmann Ivo Kabadshow, Matthias Werner, "Tasking Meets GPUs: Fighting Deadlocks and Other Monsters", 2019
@misc{morgenstern:2019:PASC,
author={Laura Morgenstern and Andreas Beckmann Ivo Kabadshow and Matthias Werner},
title={{Tasking Meets GPUs: Fighting Deadlocks and Other Monsters}},
url={https://pasc19.pasc-conference.org/program/schedule/presentation/?id=post131&sess=sess179},
howpublished={{Poster at PASC19 (Platform for Advanced Scientific Computing Conference), Zürich}},
keywords= {scalability},
year=2019
- Laura Morgenstern, David Haensel, Andreas Beckmann, Ivo Kabadshow, "NUMA-Awareness as a Plug-In", 2019
@misc{morgenstern:2019:deRSE,
author={Laura Morgenstern and David Haensel and Andreas Beckmann and Ivo Kabadshow},
title={{NUMA-Awareness as a Plug-In}},
url={https://www.researchgate.net/publication/333828565_NUMA-Awareness_as_a_Plug-In},
howpublished={{Poster at deRSE19}},
keywords= {scalability},
year=2019
- Laura Morgenstern, Andreas Beckmann, Ivo Kabadshow, "Towards Supporting Heterogeneous Hardware in Gromacs", 2018
@misc{morgenstern:2018:ISC,
author={Laura Morgenstern and Andreas Beckmann and Ivo Kabadshow},
title={{Towards Supporting Heterogeneous Hardware in Gromacs}},
url={https://www.researchgate.net/publication/326322310_Towards_Supporting_Heterogeneous_Hardware_in_Gromacs},
howpublished={{Project Poster at ISC High Performance 2018}},
month=6,
keywords= {scalability},
year=2018,
doi={10.13140/RG.2.2.27555.35365}
- Laura Morgenstern, "Scalability Enhancements to FMM for MD Simulations", 2018
@misc{morgenstern:2018:jlesc,
author={Laura Morgenstern},
title={{Scalability Enhancements to FMM for MD Simulations}},
url={https://www.researchgate.net/publication/326325786_Scalability_Enhancements_to_FMM_for_MD_Simulations},
howpublished={{Project Poster at ISC High Performance 2018}},
month=4,
keywords= {scalability},
year=2018,
doi={10.13140/RG.2.2.19166.74561}
- Laura Morgenstern, "A NUMA-aware Task-based Load-balancing Scheme for the Fast Multipole Method", 2018
@misc{morgenstern:2018:ISINA,
author={Laura Morgenstern},
title={{A NUMA-aware Task-based Load-balancing Scheme for the Fast Multipole Method}},
url={https://www.researchgate.net/publication/323573246_A_NUMA-aware_Task-based_Load-balancing_Scheme_for_the_Fast_Multipole_Method},
howpublished={{ISINA Best-Poster-Award (Interdisziplinäres Forum für Ingenieurinnen und Naturwissenschaftlerinnen)}},
month=3,
keywords= {scalability},
year=2018,
doi={10.13140/RG.2.2.22152.32009}
- Laura Morgenstern, "Towards a NUMA-aware Task-based Fast Multipole Method", in Fall School of the HPI Future SOC Lab on Efficient Architectures for Data Science (EADS 2017), 2017 @inproceedings{morgenstern:2017:NUMA,
title = {Towards a NUMA-aware Task-based Fast Multipole Method},
author = {Laura Morgenstern},
year = 2017,
booktitle = {Fall School of the HPI Future SOC Lab on Efficient Architectures for Data Science (EADS 2017)},
address = {Potsdam},
keywords= {scalability},
doi = {10.13140/RG.2.2.19764.14725}
2020
2019
2018
2017