-
Cross-platform programming model for many-core lattice Boltzmann simulations
Authors:
Jonas Latt,
Christophe Coreixas,
Joël Beny
Abstract:
We present a novel, hardware-agnostic implementation strategy for lattice Boltzmann (LB) simulations, which yields massive performance on homogeneous and heterogeneous many-core platforms. Based solely on C++17 Parallel Algorithms, our approach does not rely on any language extensions, external libraries, vendor-specific code annotations, or pre-compilation steps. Thanks in particular to a recentl…
▽ More
We present a novel, hardware-agnostic implementation strategy for lattice Boltzmann (LB) simulations, which yields massive performance on homogeneous and heterogeneous many-core platforms. Based solely on C++17 Parallel Algorithms, our approach does not rely on any language extensions, external libraries, vendor-specific code annotations, or pre-compilation steps. Thanks in particular to a recently proposed GPU back-end to C++17 Parallel Algorithms, it is shown that a single code can compile and reach state-of-the-art performance on both many-core CPU and GPU environments for the solution of a given non trivial fluid dynamics problem. The proposed strategy is tested with six different, commonly used implementation schemes to test the performance impact of memory access patterns on different platforms. Nine different LB collision models are included in the tests and exhibit good performance, demonstrating the versatility of our parallel approach. This work shows that it is less than ever necessary to draw a distinction between research and production software, as a concise and generic LB implementation yields performances comparable to those achievable in a hardware specific programming language. The results also highlight the gains of performance achieved by modern many-core CPUs and their apparent capability to narrow the gap with the traditionally massively faster GPU platforms. All code is made available to the community in form of the open-source project "stlbm", which serves both as a stand-alone simulation software and as a collection of reusable patterns for the acceleration of pre-existing LB codes.
△ Less
Submitted 9 May, 2021; v1 submitted 22 October, 2020;
originally announced October 2020.
-
Digital Blood in Massively Parallel CPU/GPU Systems for the Study of Platelet Transport
Authors:
Christos Kotsalos,
Jonas Latt,
Joel Beny,
Bastien Chopard
Abstract:
We propose a highly versatile computational framework for the simulation of cellular blood flow focusing on extreme performance without compromising accuracy or complexity. The tool couples the lattice Boltzmann solver Palabos for the simulation of the blood plasma, a novel finite element method (FEM) solver for the resolution of the deformable blood cells, and an immersed boundary method for the…
▽ More
We propose a highly versatile computational framework for the simulation of cellular blood flow focusing on extreme performance without compromising accuracy or complexity. The tool couples the lattice Boltzmann solver Palabos for the simulation of the blood plasma, a novel finite element method (FEM) solver for the resolution of the deformable blood cells, and an immersed boundary method for the coupling of the two phases. The design of the tool supports hybrid CPU-GPU executions (fluid, fluid-solid interaction on CPUs, the FEM solver on GPUs), and is non-intrusive, as each of the three components can be replaced in a modular way. The FEM-based kernel for solid dynamics outperforms other FEM solvers and its performance is comparable to the state-of-the-art mass-spring systems. We perform an exhaustive performance analysis on Piz Daint at the Swiss National Supercomputing Centre and provide case studies focused on platelet transport. The tests show that this versatile framework combines unprecedented accuracy with massive performance, rendering it suitable for the upcoming exascale architectures.
△ Less
Submitted 8 November, 2019;
originally announced November 2019.
-
Efficient supersonic flows through high-order guided equilibrium with lattice Boltzmann
Authors:
Jonas Latt,
Christophe Coreixas,
Joel Beny,
Andrea Parmigiani
Abstract:
A double-distribution-function based lattice Boltzmann method (DDF-LBM) is proposed for the simulation of polyatomic gases in the supersonic regime. The model relies on an extended equilibrium state that is constructed to reproduce the first 13 moments of the Maxwell-Boltzmann distribution exactly. This extends the validity of the standard 5-constraint (mass, momentum and energy) approach and lead…
▽ More
A double-distribution-function based lattice Boltzmann method (DDF-LBM) is proposed for the simulation of polyatomic gases in the supersonic regime. The model relies on an extended equilibrium state that is constructed to reproduce the first 13 moments of the Maxwell-Boltzmann distribution exactly. This extends the validity of the standard 5-constraint (mass, momentum and energy) approach and leads to the correct simulation of thermal, compressible flows with only 39 discrete velocities in 3D. The stability of this BGK-LBM is reinforced by relying on Knudsen-number-dependent relaxation times that are computed analytically. Hence, high-Reynolds number, supersonic flows can be simulated in an efficient and elegant manner. While the 1D Riemann problem shows the ability of the proposed approach to handle discontinuities in the zero-viscosity limit, the simulation of the flow past a NACA0012 airfoil (Mach number $\mathrm{Ma}=1.5$, Reynolds number $\mathrm{Re=10^4}$) confirms the excellent behavior of this model in a low-viscosity and supersonic regime. The proposed model is substantially more efficient than the previous 5-moment D3Q343 DDF-LBM and opens up a whole new world of compressible flow applications that can be realistically tackled with a purely LB approach.
△ Less
Submitted 12 February, 2020; v1 submitted 29 October, 2019;
originally announced October 2019.
-
Efficient LBM on GPUs for dense moving objects using immersed boundary condition
Authors:
Joel Beny,
Jonas Latt
Abstract:
There exists an increasing interest for using immersed boundary methods (IBMs) (Peskin 2000) to model moving objects in computational fluid dynamics. Indeed, this approach is particularly efficient, because the fluid mesh does not require to be body-fitted or to adjust dynamically to the motion of the body. Frequently, IBMs are implemented in combination with the lattice Boltzmann methods (LBM) (K…
▽ More
There exists an increasing interest for using immersed boundary methods (IBMs) (Peskin 2000) to model moving objects in computational fluid dynamics. Indeed, this approach is particularly efficient, because the fluid mesh does not require to be body-fitted or to adjust dynamically to the motion of the body. Frequently, IBMs are implemented in combination with the lattice Boltzmann methods (LBM) (Krüger 2016). They fit elegantly into the framework of this method, and yield impressive parallel performances. It has also become quite common to accelerate LBM simulations with the use of Graphics Processing Units (GPUs) (Tölke 2010), as the underlying algorithm adjusts naturally to the architecture of such platforms. It is not uncommon that speedups of an order of magnitude, or more, at equal financial cost or energy consumption are observed, as compared to classical CPUs. IBM algorithms are however more difficult to adapt to GPUs, because their complex memory access pattern conflicts with a GPU's strategy of broadcasting data to a large number of GPU cores in single memory accesses. In the existing literature, GPU implementations of LBM-IBM codes are therefore restricted to situations in which the immersed surfaces are very small compared to the total number of fluid cells (Valero-Lara 2014), as is often the case in exterior flow simulations around an obstacle. This assumption is however not valid in many other cases of interest.
We propose a new method for the implementation of a LBM-IBM on GPUs in the CUDA language, which allows to handle a substantially larger immersed surfaces with acceptable performance than previous implementations.
△ Less
Submitted 27 March, 2019;
originally announced April 2019.
-
Integration of iron in natural and synthetic Al-pyrophyllites: an infrared spectroscopic study
Authors:
Sébastien Lantenois,
Jean-Michel Bény,
Fabrice Muller,
Rémi Champallier
Abstract:
Numerous studies focus on the relationships between chemical composition and OHband positions in the infrared (IR) spectra of micaceous minerals. These studies are based on the coexistence, in dioctahedral micas or smectites, of several cationic pairs around the hydroxyl group which each produce a characteristic band in the IR spectrum. The aim of this work is to obtain the wavenumber values of…
▽ More
Numerous studies focus on the relationships between chemical composition and OHband positions in the infrared (IR) spectra of micaceous minerals. These studies are based on the coexistence, in dioctahedral micas or smectites, of several cationic pairs around the hydroxyl group which each produce a characteristic band in the IR spectrum. The aim of this work is to obtain the wavenumber values of the IR OH vibration bands of the (Al-Fe3+)-OH and (Fe3+-Fe3+)-OH local cationic environments of 'pyrophyllite type' in order to prove, disprove or modify a model of dioctahedral phyllosilicate OH-stretching band decomposition. Natural samples are characterized by powder X-ray diffraction (XRD), Fourier transform infrared (FTIR) and Raman spectroscopies and electron microprobe; the hydrothermal synthesis products are also analysed by powder XRD and FTIR after inductively coupled plasma measurements to obtain the chemical compositions of nascent gel phases. Natural samples contain some impurities which were eliminated after acid treatment; nevertheless, a small Fe content is found in the pyrophyllite structure. The amount of Fe which is incorporated within the pyrophyllite structure is much more important for the synthetic samples than for the natural ones. The IR OH bands were clearly observed in both natural and synthetic pyrophyllites and assigned to hydroxides bonded to (Al-Al), (Al-Fe) and (Fe-Fe) cationic pairs. During this study, three samples were analysed by DTG to check the cis- or trans-vacant character of the layers and to determine the influence of this structural character on the OH-stretching band position in IR spectroscopy.
△ Less
Submitted 10 July, 2007;
originally announced July 2007.
-
Incorporation of H_2 in vitreous silica, qualitative and quantitative determination from Raman and infrared spectroscopy
Authors:
Burkhard C. Schmidt,
François Holtz,
Jean-Michel Bény
Abstract:
Incorporation mechanisms of H_2 in silica glass were studied with Raman and infrared (IR) microspectroscopy. Hydrogenated samples were prepared at temperatures between 800 deg C and 955 deg C at 2 kbar total pressure. Hydrogen fugacities (f_{H_2}) were controlled using the double capsule technique with the iron-wüstite (IW) buffer assemblage generating f_{H_2} of 1290-1370 bars corresponding to…
▽ More
Incorporation mechanisms of H_2 in silica glass were studied with Raman and infrared (IR) microspectroscopy. Hydrogenated samples were prepared at temperatures between 800 deg C and 955 deg C at 2 kbar total pressure. Hydrogen fugacities (f_{H_2}) were controlled using the double capsule technique with the iron-wüstite (IW) buffer assemblage generating f_{H_2} of 1290-1370 bars corresponding to H_2 partial pressures (P_{H_2}) of 960-975 bars. We found that silica glass hydrogenated under such conditions contains molecular hydrogen (H_2) in addition to SiH and SiOH groups. H_2 molecules dissolved in the quenched glasses introduce a band at 4136 cm^{-1} in the Raman spectra which in comparison to that of gaseous H_2 is wider and is shifted to lower frequency. IR spectra of hydrogenated samples contain a band at 4138 cm^{-1} which we assign to the stretching vibration of H_2 molecules located in non-centrosymmetric sites. The Raman and IR spectra indicate that the dissolved H_2 molecules interact with the silicate network. We suggest that the H_2 band is the envelope of at least three components due to the occupation of at least three different interstitial sites by H_2 molecules. Both, Raman and IR spectra of hydrogenated glasses contain bands at ~2255 cm^{-1} which may be due to the vibration of SiH groups.
△ Less
Submitted 31 August, 2006;
originally announced August 2006.
