Statement

HPC clusters are indispensable for the solution of real-world large-scale engineering problems. Knowing potential scalability limitations of current standard processors, different many-core architectures have been proposed. They are a prototype for hardware in future HPC clusters. An important research objective is the introduction of new hardware-aware numerical methods that can profit from these new technologies. This also includes resilience considerations.

Exemplary results

Multi-GPU parallelization of a two-phase Navier-Stokes solver
Power consumption of the two-phase flow solver NaSt3DGPF comparing GPU and CPU implementations. =

Power consumption of the two-phase flow solver NaSt3DGPF comparing GPU and CPU implementations. =

Weak scaling parallel efficiency of the multi-GPU parallel Navier-Stokes solver NaSt3DGPF

Weak scaling parallel efficiency of the multi-GPU parallel Navier-Stokes solver NaSt3DGPF

Related Work

  • H. Harbrecht and P. Zaspel. A scalable H-matrix approach for the solution of boundary integral equations on multi-GPU clusters. Preprint 2018-11, Fachbereich Mathematik, Universität Basel, Switzerland, 2018. Also available as arXiv:1806.11558.

  • P. Zaspel. Algorithmic patterns for H matrices on many-core processors, accepted for publication in Journal of Scientific Computing, Springer, August 2018. Also available as Preprint 2017-12, Fachbereich Mathematik, Universität Basel, Switzerland, 2017 and as arXiv:1708.09707 preprint.

  • P. Zaspel. Analysis and parallelization strategies for Ruge-Stüben AMG on many-core processors, Preprint 2017-06, Fachbereich Mathematik, Universität Basel, Switzerland, 2017.

  • P. Zaspel. Parallel RBF Kernel-Based Stochastic Collocation for Large-Scale Random PDEs, PhD Thesis, Institute for Numerical Simulation, University of Bonn, Germany, Apr. 2015

  • EXAHD: An Exa-scalable Two-Level Sparse Grid Approach for Higher-Dimensional Problems in Plasma Physics and Beyond. In: Lopes L. et al. (eds) Euro-Par 2014: Parallel Processing Workshops. Euro-Par 2014. Lecture Notes in Computer Science, vol 8806. Springer, Cham, 2014.

  • P. Zaspel and M. Griebel. Solving incompressible two-phase flows on multi-GPU clusters. Computer & Fluids, 80(0):356 - 364, 2013.

  • P. Zaspel and M. Griebel. Massively parallel fluid simulations on amazon's hpc cloud. In Network Cloud Computing and Applications (NCCA), 2011 First International Symposium on, pages 73 -78, Nov. 2011.

  • M. Griebel and P. Zaspel. A multi-GPU accelerated solver for the three-dimensional two-phase incompressible Navier-Stokes equations. Computer Science - Research and Development, 25(1-2):65-73, May 2010.

  • V. Heuveline, M. Schick, C. Webster, P. Zaspel. Uncertainty Quantification and High Performance Computing, Dagstuhl Reports, Vol. 6, Issue 9, pp. 59-73.

 Software

  • hmglib: hierarchical matrices on graphic processing units (github)

  • MPLA: multi-GPU parallel library for dense iterative matrix solvers (github)

  • Multi-GPU support und uncertainty quantification for two-phase Navier Stokes (NaSt3DGPF)