Load-Balanced Local Time Stepping for Large-Scale Wave Propagation

Rietmann, Max; Peter, Daniel; Schenk, Olaf; Uçar, Bora; Grote, Marcus J.

doi:10.1109/ipdps.2015.10

Cited by 20 publications

(13 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We believe that our method will be more efficient than implicit schemes on a larger number of CPU cores, thanks to MPI or hybrid MPI/OpenMP parallelization. This requires a specific work on load balancing for our local time stepping scheme, like in [36] and [37]. This is a work in progress that will be the subject of a future publication.…”

Section: Resultsmentioning

confidence: 99%

On some explicit local time stepping finite volume schemes for CFD

Jeanmasson

Mary

Mieussens

2019

Journal of Computational Physics

View full text Add to dashboard Cite

In this paper, two local time stepping schemes of order two and three in time are proposed. By construction, they are not mass conservative but a correction stage is added to make them conservative. These schemes are compared with some local time stepping schemes existing in the literature (schemes of Constantinescu and Sandu). The comparisons are carried out on various test cases. They prove that our schemes are efficient and our third order local time stepping has a higher time accuracy than the schemes based on the strategy of Constantinescu and Sandu. Our third order local time stepping scheme is used to perform an industrial-like test case: a 3D Large-Eddy Simulation over an airfoil.

show abstract

Section: Resultsmentioning

confidence: 99%

On some explicit local time stepping finite volume schemes for CFD

Jeanmasson

Mary

Mieussens

2019

Journal of Computational Physics

View full text Add to dashboard Cite

show abstract

“…Moving the SEM to a GPU architecture can result in significant speedups depending on the setup (Komatitsch et al 2010;Rietmann et al 2012;. The implementation of local timestepping (Dumbser et al 2007;Rietmann et al 2015) can result in similar speedups for specific problems where a relatively low number of small elements limit the global timestep of the simulation. Simulation cost has also been reduced with simplifications of the full 3-D problem.…”

Section: Reducing the Computational Costmentioning

confidence: 99%

Accelerating Numerical Wave Propagation by Wavefield Adapted Meshes, Part II: Full-Waveform Inversion

Thrastarson¹,

Driel²,

Krischer³

et al. 2019

Preprint

View full text Add to dashboard Cite

We present a novel full-waveform inversion approach which can reduce the computational cost by up to an order of magnitude compared to conventional approaches, provided that variations in medium properties are sufficiently smooth. Our method is based on the usage of wavefield-adapted meshes which accelerate the forward and adjoint wavefield simulations. By adapting the mesh to the expected complexity and smoothness of the wavefield, the number of elements needed to discretise the wave equation can be greatly reduced. This leads to spectral-element meshes which are optimally tailored to source locations and medium complexity. We demonstrate a workflow which opens up the possibility to use these meshes in full-waveform inversion and show the computational advantages of the approach. We provide examples in 2-D and 3-D to illustrate the concept, describe how the new workflow deviates from the standard full-waveform inversion workflow, and explain the additional steps in detail.

show abstract

“…A more detailed analysis of the parallelization of our LTS-Newmark method is beyond the scope of this article, but is examined and compared with several parallelization methods in a concurrent article [35].…”

Section: Load Balancingmentioning

confidence: 99%

“…In fact, it remains our partitioning scheme of choice, performing as well or better than the multi-constrain approach that uses the PaToH partitioning library, however the difference was small. The examples shown in the next section's performance experiments used the simpler approach, however for more details and a comparison of the two schemes, see our concurrent paper focused entirely on the partitioning problem [35].…”

Section: Load Balancingmentioning

confidence: 99%

Newmark local time stepping on high-performance computing architectures

Rietmann

Grote

Peter

et al. 2017

Journal of Computational Physics

Self Cite

View full text Add to dashboard Cite

Please cite this article in press as: M. Rietmann et al., Newmark local time stepping on high-performance computing architectures, J. Comput. Phys. (2016), http://dx. AbstractIn multi-scale complex media, finite element meshes often require areas of local refinement, creating small elements that can dramatically reduce the global time-step for wave-propagation problems due to the CFL condition. Local time stepping (LTS) algorithms allow an explicit time-stepping scheme to adapt the time-step to the element size, allowing near-optimal time-steps everywhere in the mesh. We develop an efficient multilevel LTS-Newmark scheme and implement it in a widely used continuous finite element seismic wave-propagation package. In particular, we extend the standard LTS formulation with adaptations to continuous finite element methods that can be implemented very efficiently with very strong element-size contrasts (more than 100x). Capable of running on large CPU and GPU clusters, we present both synthetic validation examples and large scale, realistic application examples to demonstrate the performance and applicability of the method and implementation on thousands of CPU cores and hundreds of GPUs.

show abstract

Load-Balanced Local Time Stepping for Large-Scale Wave Propagation

Cited by 20 publications

References 10 publications

On some explicit local time stepping finite volume schemes for CFD

On some explicit local time stepping finite volume schemes for CFD

Accelerating Numerical Wave Propagation by Wavefield Adapted Meshes, Part II: Full-Waveform Inversion

Newmark local time stepping on high-performance computing architectures

Contact Info

Product

Resources

About