Layout fehlt: setlanguage

SeASiTe – Selbstadaption für zeitschrittbasierte Simulationstechniken auf heterogenen HPC-Systemen

Projektstart: 01.01.2017, Projektende: 31.12.2019

Geldgeber:

BMBF

Projektbeteiligte

Projektleiter

Prof. Dr. Thomas Rauber, PD Dr. Matthias Korch

Ansprechpartner

PD Dr. Matthias Korch

Projektmitarbeiter

Johannes Seiferth, M.Sc.

Markus Scherg, M.Sc.

Externe Partner

Projektbeschreibung

Das Forschungsprojekt SeASiTe stellt sich der Aufgabe, eine systematische Untersuchung von Selbstadaption für zeitschrittbasierte Simulationstechniken auf heterogenen HPC-Systemen durchzuführen. Das Ziel ist der Entwurf und die Bereitstellung des Prototypen eines Werkzeugkastens, mit dessen Hilfe Programmierer ihre Anwendungen mit effizienten Selbstadaptionstechniken ausstatten können. Der Ansatz beinhaltet die Selbstadaption sowohl hinsichtlich relevanter System- und Programmparameter als auch möglicher Programmtransformationen. Die Optimierung der Programmausführung für mehrere nicht-funktionale Ziele (z.B. Laufzeit oder Energieverbrauch) soll auf einer Performance-Modellierung zur Eingrenzung des Suchraums effizienter Programmvarianten aufbauen. Anwendungsunabhängige Methoden und Strategien zur Selbstadaption sollen in einem Autotuning-Navigator gekapselt werden.

Das Verbundprojekt "SeASiTe" ist ein Forschungsvorhaben auf dem Gebiet "Grundlagenorientierte Forschung für HPC-Software im Hoch- und Höchstleistungsrechnen" und wird im Rahmen des Förderprogramms "IKT 2020 – Forschung für Innovationen" vom Bundesministerium für Bildung und Forschung (BMBF) gefördert.

Veröffentlichungen

2019

  • T. Jakobs, B. Naumann and G. Rünger.
    Performance and energy consumption of the SIMD Gram–Schmidt process for vector orthogonalization.
    The Journal of Supercomputing.
    Springer, 2019.
    DOI: 10.1007/s11227-019-02839-0
  • S. Margenov, T. Rauber, E. Atanassov, F. Almeida, V. Blanco, R. Ciegis, A. Cabrera, N. Frasheri, S. Harizanov, R. Kriauzien, G. Rünger, P. S. Segundo, A. Starikovicius, S. Szabo and B. Zavalnij.
    Applications for ultrascale systems.
    In Ultrascale Computing Systems, IET Professional Applications of Computing 24, pages 189–244.
    The Institution of Engineering and Technology, 2019.
    DOI: 10.1049/PBPC024E_ch6
  • T. Rauber and G. Rünger.
    DVFS RK: Performance and Energy Modeling of Frequency-Scaled Multi-threaded Runge-Kutta Methods.
    In Proceedings of the 27th Euromicro International Conference on Parallel, Distributed and Network-based Processing (PDP 2019), pages 392–399.
    IEEE, 2019.
    DOI: 10.1109/EMPDP.2019.8671593
  • T. Rauber and G. Rünger.
    Enabling Scalability, Adaptivity, and Resilience in Cloud Applications by Software-defined M-Task-based Programming.
    In Proceedings of the 6th International Conference on Software Defined Systems (SDS 2019), pages 88–95.
    IEEE, 2019.
    DOI: 10.1109/SDS.2019.8768599
  • T. Rauber, G. Rünger and M. Stachowski.
    Model-based optimization of the energy efficiency of multi-threaded applications.
    Sustainable Computing: Informatics and Systems, 22:44–61, 2019.
    Elsevier, 2019.
    DOI: 10.1016/j.suscom.2019.01.022
  • M. Scherg, J. Seiferth, M. Korch and T. Rauber.
    Performance Prediction of Explicit ODE Methods on Multi-Core Cluster Systems.
    In Proceedings of the 2019 ACM/SPEC International Conference on Performance Engineering (ICPE ’19), pages 139–150.
    ACM, 2019.
    DOI: 10.1145/3297663.3310306
  • 2018

    • G. Hager, J. Eitzinger, J. Hornich, F. Cremonesi, C. L. Alappat, T. Röhl and G. Wellein.
      Applying the Execution-Cache-Memory Model: Current State of Practice.
      In The International Conference for High Performance Computing, Networking, Storage, and Analysis (SC 2018).
      Poster.
    • J. Hofmann, G. Hager, and D. Fey.
      On the accuracy and usefulness of analytic energy models for contemporary multicore processors.
      In High Performance Computing. ISC High Performance 2018, LNCS 10876, pages 22–43.
      Springer, 2018.
      Gauss Award (Press Release).
      DOI: 10.1007/978-3-319-92040-5_2
    • M. Hofmann, R. Kiesel, D. Leichsenring, and G. Rünger.
      A hybrid CPU/GPU implementation of computationally intensive particle simulations using OpenCL.
      To appear in Proceedings of the 17th IEEE International Symposium On Parallel And Distributed Computing (ISPDC 2018)., pages 9–16.
      IEEE, 2018.
      DOI: 10.1109/ISPDC2018.2018.00011
    • M. Hofmann, R. Kiesel, and G. Rünger.
      Energy and Performance Analysis of Parallel Particle Solvers from the ScaFaCoS Library.
      In Proceedings of the 2018 ACM/SPEC International Conference on Performance Engineering (ICPE 2018), pages 88–95.
      ACM, 2018.
      DOI: 10.1145/3184407.3184409
    • M. Hofmann and G. Rünger.
      Flexible all-to-all data redistribution methods for grid-based particle codes.
      Concurrency and Computation: Practice and Experience, volume 30.
      Wiley, 2018.
      DOI: 10.1002/cpe.4421
    • N. Kalinnik, R. Kiesel, T. Rauber, M. Richter, and G. Rünger.
      Exploring Self-Adaptivity towards Performance and Energy for Time-stepping Methods.
      In Proceedings of the 30th International Symposium on Computer Architecture and High Performance Computing (SBAC-PAD 2018), pages 115–123.
      IEEE, 2018.
      DOI: 10.1109/CAHPC.2018.8645887
    • N. Kalinnik, R. Kiesel, T. Rauber, M. Richter, and G. Rünger.
      On the Autotuning Potential of Time-stepping methods from Scientific Computing.
      In Proceedings of the 2018 Federated Conference on Computer Science and Information Systems (FedCSIS 2018), 11th Workshop on Computer Aspects of Numerical Algorithms (CANA'18), ACSIS 15, pages 329–338.
      IEEE, 2018.
      DOI: 10.15439/2018F169
    • T. Rauber and G. Rünger.
      A Scheduling Selection Process for Energy-Efficient Task Execution on DVFS Processors.
      Concurrency and Computation: Practice and Experience.
      Special Issue Paper.
      Wiley, 2018.
      DOI: 10.1002/cpe.5043
    • T. Rauber and G. Rünger.
      Comparison of Time and Energy Oriented Scheduling for Task-Based Programs.
      In Parallel Processing and Applied Mathematics. PPAM 2017, LNCS 10777, pages 185–196.
      Springer, 2018.
      DOI: 10.1007/978-3-319-78024-5_17
    • T. Rauber and G. Rünger.
      How do loop transformations affect the energy consumption of Runge-Kutta methods?
      In Proceedings of the 26th Euromicro International Conference on Parallel, Distributed, and Network-based Processing (PDP 2018), pages 499–507.
      IEEE, 2018.
      DOI: 10.1109/PDP2018.2018.00085
    • T. Rauber and G. Rünger.
      Energy and Performance Improvement of Parallel ODE Solvers by Application-specific Program Transformations.
      To appear in Proceedings of the 19th IEEE International Workshop on Parallel and Distributed Scientific and Engineering Computing (PDSEC-18).
      IEEE, 2018.
    • M. Richter and G. Rünger.
      Symbolic matrix multiplication for multi-threaded sparse GEMM utilizing sparse matrix formats.
      In International Conference on High Performance Computing & Simulation (HPCS 2018), pages 523–530.
      IEEE, 2018.
      DOI: 10.1109/HPCS.2018.00088
    • J. Seiferth, C. Alappat, M. Korch, and T. Rauber.
      Applicability of the ECM Performance Model to Explicit ODE Methods on Current Multi-Core Processors.
      In High Performance Computing. ISC High Performance 2018, LNCS 10876, pages 163–183.
      Springer, 2018.
      DOI: 10.1007/978-3-319-92040-5_9

    2017

    • T. Rauber and G. Rünger.
      Tuning Energy Effort and Execution Time of Application Software.
      In Information Systems Architecture and Technology: Proceedings of 38th International Conference on Information Systems Architecture and Technology – ISAT 2017, AISC 656, pages 239–251.
      Springer, 2017.
      DOI: 10.1007/978-3-319-67229-8_22
    • T. Rauber, G. Rünger, and M. Stachowski.
      Model-based Optimization of the Energy Efficiency of Multi-threaded Applications.
      In Proceedings of the 8th International Green Sustainable Computing Conference (IGSC’17).
      IEEE, 2017.
      DOI: 10.1109/IGCC.2017.8323578
    • T. Rauber, G. Rünger, and M. Stachowski.
      Performance and Energy Metrics for Multi-threaded Applications on DVFS Processors.
      Sustainable Computing: Informatics and Systems, volume 17, pages 55–68.
      Elsevier, 2017.
      DOI: 10.1016/j.suscom.2017.10.015
    • T. Rauber, G. Rünger, and M. Stachowski.
      Towards New Metrics for Appraising Performance and Energy Efficiency of Parallel Scientific Programs.
      In Proceedings of the 13th IEEE International Conference on Green Computing and Communication (GreenCom-2017).
      IEEE, 2017. DOI: 10.1109/iThings-GreenCom-CPSCom-SmartData.2017.75

    Universität Bayreuth -