Multi-domain Grid Refinement for Lattice-Boltzmann Simulations on Heterogeneous Platforms

Valero-Lara, Pedro; Jansson, Johan

doi:10.1109/cse.2015.9

Cited by 8 publications

(9 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…These methodologies have been analyzed deeply and validated in several numerical scenarios , so we focus on the implementation techniques adopted to keep the solver highly efficient on CPU+GPU heterogeneous platforms. The present work extends the previously published work with additional contributions. In particular, this work includes a comparative study (in terms of performance and numerical accuracy) among two different refinement methods for LBM simulations.…”

Section: Introductionsupporting

confidence: 82%

“…As mentioned earlier, the equation is typically advanced in time in two stages, the collision and the streaming stages . Given f i ( x , t ) compute:

\begin{matrix} ρ & = big\sum f_{i} (boldx, t) and \\ ρ boldu & = big\sum {bolde}_{i} f_{i} (boldx, t) \end{matrix}

Collision stage:

f_{i}^{*} (x, t + Δ t) = f_{i} (x, t) - \frac{Δ t}{τ} (f (x, t) - f_{i}^{eq} (x, t))

Streaming stage:

f_{i} (x + e_{i} Δ t, t + Δ t) = f_{i}^{*} (x, t + Δ t)

…”

Section: Lattice‐boltzmann Methodsmentioning

confidence: 99%

“…As mentioned earlier, the equation 6 is typically advanced in time in two stages, the collision and the streaming stages [11,23,38]. Given f i .x; t/ compute:…”

Section: Of 20mentioning

confidence: 99%

See 2 more Smart Citations

Heterogeneous CPU+GPU approaches for mesh refinement over Lattice‐Boltzmann simulations

Valero-Lara

Jansson

2016

Concurrency and Computation

Self Cite

View full text Add to dashboard Cite

Summary The use of mesh refinement in CFD is an efficient and widely used methodology to minimize the computational cost by solving those regions of high geometrical complexity with a finer grid. In this work, the author focuses on studying two methods, one based on Multi‐Domain and one based on Irregular meshing, to deal with mesh refinement over LBM simulations. The numerical formulation is presented in detail. It is proposed two approaches, homogeneous GPU and heterogeneous CPU+GPU, on each of the refinement methods. Obviously, the use of the two architectures, CPU and GPU, to compute the same problem involves more important challenges with respect to the homogeneous counterpart. These challenges and the strategies to deal with them are described in detail into the present work. We pay a particular attention to the differences among both methodologies/implementations in terms of programmability, memory management, and performance. The size of the refined sub‐domain has important consequences over both methodologies; however, the influence on Multi‐Domain approach is much higher. For instance, when dealing with a big refined sub‐domain, the Multi‐Domain approach achieves an important fall in performance with respect to other cases, where the size of the refined sub‐domain is smaller. Otherwise, using the Irregular approach, there is no such a dramatic fall in performance when increasing the size of the refined sub‐domain. Copyright © 2016 John Wiley & Sons, Ltd.

show abstract

Section: Introductionsupporting

confidence: 82%

“…As mentioned earlier, the equation is typically advanced in time in two stages, the collision and the streaming stages . Given f i ( x , t ) compute:

\begin{matrix} ρ & = big\sum f_{i} (boldx, t) and \\ ρ boldu & = big\sum {bolde}_{i} f_{i} (boldx, t) \end{matrix}

Collision stage:

f_{i}^{*} (x, t + Δ t) = f_{i} (x, t) - \frac{Δ t}{τ} (f (x, t) - f_{i}^{eq} (x, t))

Streaming stage:

f_{i} (x + e_{i} Δ t, t + Δ t) = f_{i}^{*} (x, t + Δ t)

…”

Section: Lattice‐boltzmann Methodsmentioning

confidence: 99%

See 1 more Smart Citation

Heterogeneous CPU+GPU approaches for mesh refinement over Lattice‐Boltzmann simulations

Valero-Lara

Jansson

2016

Concurrency and Computation

Self Cite

View full text Add to dashboard Cite

show abstract

“…Unlike the previous strategy, this approach ( LBM‐Swap ) does not need more memory ( ghost cells ) or change the data layout. This makes much easier the implementation and the integration with the CPU for heterogeneous implementations . The LBM‐Swap algorithm only needs one lattice‐speed data space.…”

Section: Lbm‐swapmentioning

confidence: 99%

Reducing memory requirements for large size LBM simulations on GPUs

Valero-Lara

2017

Concurrency and Computation

Self Cite

View full text Add to dashboard Cite

Summary The scientific community in its never‐ending road of larger and more efficient computational resources is in need of more efficient implementations that can adapt efficiently on the current parallel platforms. Graphics processing units are an appropriate platform that cover some of these demands. This architecture presents a high performance with a reduced cost and an efficient power consumption. However, the memory capacity in these devices is reduced and so expensive memory transfers are necessary to deal with big problems. Today, the lattice‐Boltzmann method (LBM) has positioned as an efficient approach for Computational Fluid Dynamics simulations. Despite this method is particularly amenable to be efficiently parallelized, it is in need of a considerable memory capacity, which is the consequence of a dramatic fall in performance when dealing with large simulations. In this work, we propose some initiatives to minimize such demand of memory, which allows us to execute bigger simulations on the same platform without additional memory transfers, keeping a high performance. In particular, we present 2 new implementations, LBM‐Ghost and LBM‐Swap, which are deeply analyzed, presenting the pros and cons of each of them.

show abstract

“…Such is the case of [15], where a Lattice Boltzman solver was implemented on GPUs to solve the pollution dispersion in an urban environment. Moreover, due to the high data parallelism found in the Lattice Boltzman Method, further studies have been conducted implementing different models in the GPUs, as in the cases of Fluid-Solid interaction [16][17][18] and multi-domain grid refinement [19], to name a few, always achieving time improvements when using GPUs.…”

Section: Introductionmentioning

confidence: 99%

Heterogeneous Computing (CPU–GPU) for Pollution Dispersion in an Urban Environment

2020

View full text Add to dashboard Cite

The use of Computational Fluid Dynamics (CFD) to assist in air quality studies in urban environments can provide accurate results for the dispersion of pollutants. However, due to the computational resources needed, simulation domain sizes tend to be limited. This study aims to improve the computational efficiency of an emission and dispersion model implemented in a CPU-based solver by migrating it to a CPU–GPU-based one. The migration of the functions that handle boundary conditions and source terms for the pollutants is explained, as well as the main differences present in the solvers used. Once implemented, the model was used to run simulations with both engines on different platforms, enabling the comparison between them and reaching promising time improvements in favor of the use of GPUs.

show abstract

Multi-domain Grid Refinement for Lattice-Boltzmann Simulations on Heterogeneous Platforms

Cited by 8 publications

References 18 publications

Heterogeneous CPU+GPU approaches for mesh refinement over Lattice‐Boltzmann simulations

Heterogeneous CPU+GPU approaches for mesh refinement over Lattice‐Boltzmann simulations

Reducing memory requirements for large size LBM simulations on GPUs

Heterogeneous Computing (CPU–GPU) for Pollution Dispersion in an Urban Environment

Contact Info

Product

Resources

About