Intranet
cxHPC: Center of Excellence in High Performance Computing
Project summary
The efficient use of modern supercomputers is closely connected with an extensive and qualified user training and user support. Thus, the aim of the project cxHPC (formerly known as BAUWIHR) is to provide a center of excellence for scientific supercomputing at the Friedrich-Alexander University of Erlangen-Nuremberg.
cxHPC addresses the following tasks:
- HPC support: Programming, Debugging, Optimization of HPC codes. Provide test accounts to all major german supercomputer centers for testing and optimization
- HPC training: Organize HPC tutorials and HPC lectures for students and users
- HPC information: Present information about the latest development in HPC: new architectures, programming models, available resources, global trends,....
KONWIHR funding and follow-up projects
- initial KONWIHR funding: 01/2000 - 12/2003
- follow-up funding: onging funding through
STMWFK and FAU.
- OMI4papps continues the work of cxHPC with a special focus on highly parallel applications
Contact:
- Dr. Gerhard Wellein, Regionales Rechenzentrum Erlangen
Project staff:
- Jan Treibig (RRZE)
- Dr. Volker Weinberg (LRZ)
Selected publications
- T. Zeiser, G. Hager, G. Wellein: Vector computers in a world of commodity clusters, massively parallel systems and many-core many-threaded CPUs: Recent experience based on an advanced lattice Boltzmann flow solver, High Performance Computing in Science and Engineering '08: Transactions of the High Performance Computing Center, Stuttgart (HLRS) 2008 (Ed: W. E. Nagel, D. B. Kröner, M. Resch), Springer (Berlin, Heidelberg), 5 (2009) 333-347. ISBN 978-3-540-88301-2; DOI: 10.1007/978-3-540-88303-6_24
- T. Zeiser, G. Hager, G. Wellein: Vector computers in a world of commodity clusters, massively parallel systems and many-core many-threaded CPUs: Recent experience based on an advanced lattice Boltzmann flow solver, High Performance Computing in Science and Engineering '08: Transactions of the High Performance Computing Center, Stuttgart (HLRS) 2008 (Ed: W. E. Nagel, D. B. Kröner, M. Resch), Springer (Berlin, Heidelberg), 5 (2009) 333-347. ISBN 978-3-540-88301-2; DOI: 10.1007/978-3-540-88303-6_24
- S. Donath, K. Iglberger, G. Wellein, T. Zeiser, A. Nitsure, U. Rüde: Performance comparison of different parallel lattice Boltzmann implementations on multi-core multi-socket systems, International Journal of Computational Science and Engineering, 4:(1) (2008) 3-11. doi:10.1504/IJCSE.2008.021107
- G. Hager, T. Zeiser, G. Wellein: Data access optimizations for highly threaded multi-core CPUs with multiple memory controllers, Workshop on Large Scale Parallel Processing 2008 (IPDPS 2008), Miami, FL, April 18, 2008, (2008). arXiv:0712.2302
- T. Zeiser, G. Wellein, A. Nitsure, K. Iglberger, G. Hager: Introducing a parallel cache oblivious blocking approach for the lattice Boltzmann method, Progress in Computational Fluid Dynamics, 8:(1-4) (2008) 179-188. DOI: 10.1504/PCFD.2008.018088
- T. Zeiser: Simulation und Analyse von durchströmten Kugelschüttungen in engen Rohren unter Verwendung von Hochleistungsrechnern, Dissertation, Technische Fakultät, Universität Erlangen-Nürnberg, Verlag Dr. Hut (München), (2008). E-Diss
- T. Zeiser: Simulation und Analyse von durchströmten Kugelschüttungen in engen Rohren unter Verwendung von Hochleistungsrechnern, Dissertation, Technische Fakultät, Universität Erlangen-Nürnberg, Verlag Dr. Hut (München), (2008). E-Diss
- B. Bergen, G. Wellein, F. Hülsemann, U. Rüde: Hierarchical hybrid grids: achieving TERAFLOP performance on large scale finite element simulations, International Journal of Parallel, Emergent and Distributed Systems, 22:(4) (2007) 311-329. DOI: 10.1080/17445760701442218
- B. Bergen, G. Wellein, F. Hülsemann, U. Rüde: Hierarchical hybrid grids: achieving TERAFLOP performance on large scale finite element simulations, International Journal of Parallel, Emergent and Distributed Systems, 22:(4) (2007) 311-329. DOI: 10.1080/17445760701442218
- G. Wellein, T. Zeiser, G. Hager, S. Donath: On the single processor performance of simple lattice Boltzmann kernels, Computers & Fluids, 35 (2006) 910-919. DOI: 10.1016/j.compfluid.2005.02.008
- G. Hager: A Parallelized Density Matrix Renormalization Group Algorithm and its application to strongly correlated quantum systems, Dissertation, Mathematisch-Naturwissenschaftliche Fakultät, Universität Greifswald, (2005). E-Diss
- C. Körner, T. Pohl, U. Rüde, N. Thürey, T. Zeiser: Parallel lattice Boltzmann methods for CFD applications, Numerical Solution of Partial Differential Equations on Parallel Computers (Ed: A.M. Bruaset, A. Tveito), Springer (Berlin, Heidelberg), LNCSE 51 (2005) 439-465. DOI: 10.1007/3-540-31619-1_13
- G. Wellein, T. Zeiser, G. Hager, S. Donath: Single processor performance of simple lattice Boltzmann kernels, High Performance Computing in Science and Engineering, Munich 2002 - Transactions of the First Joint HLRB and KONWIHR Status and Result Workshop, October 10-11, 2002, Technical University of Munich, Germany (Ed: S. Wagner, W. Hanke, A. Bode, F. Durst), Springer (Berlin, Heidelberg), (2003), pp. 910-919.
Selected presentations and teaching
- annual lecture of RRZE at the University of Erlangen-Nuremberg: Programming techniques for supercomputers (PTfS)
- regular joint LRZ/RRZE block courses on High Performance Computing: more details (including slides from previous courses and announcements of future courses are available on the LRZ web page
- invited ICMMES talk 2006: G. Wellein, T. Zeiser, G. Hager, A. Nitsure, K. Iglberger, U. Rüde:
Introducing a cache-oblivious blocking approach for the lattice Boltzmann method, Hampton, Virginia, USA. - ICMMES tutorial 2006: G. Wellein, T. Zeiser, G. Hager, S. Donath: Efficient implementation of simple lattice Boltzmann kernels, Hampton, Virginia, USA.
- ParCFD tutorial 2007: G. Wellein, T. Zeiser, G. Hager, S. Donath: Efficient implementation of simple lattice Boltzmann kernels (Part I+II), Antalya, Turkey, May 2007.
- ICMMES tutorial 2007: T. Zeiser, G. Wellein, G. Hager: Efficient implementation of lattice Boltzmann flow solvers, Munich.
- invited ASIM08 talk: G. Wellein, G. Hager, T. Zeiser, M. Meier: Architecture and performance of multi-/many core systems, Erlangen.
- NUMET 2008: G. Hager: Effiziente Nutzung von Hochleistungsrechnern in der numerischen Strömungsmechanik, LSTM-Erlangen, March 2008.
- NUMET 2008: T. Zeiser: Strömungssimulation mit Lattice-Boltzmann-Verfahren, LSTM-Erlangen, March 2008.
- ParCFD Minisymposium New computer architectures: New potentials and new challenges for numerical simulations: T. Zeiser, G. Hager, G. Wellein: A first view to a homogeneous many-core system on a chip architecture, Lyon, France, May 2008.