Search

The lattice Boltzmann method (LBM) has emerged as a prominent technique for solving fluid dynamics problems due to its algorithmic potential for computational scalability. We introduce XLB library, a Python-based differentiable LBM library based on the JAX platform. The architecture of XLB is predicated upon ensuring accessibility, extensibility, and computational performance, enabling scaling effectively across CPU, TPU, multi-GPU, and distributed multi-GPU or TPU systems. The library can be readily augmented with novel boundary conditions, collision models, or multi-physics simulation capabilities. XLB's differentiability and data structure is compatible with the extensive JAX-based machine learning ecosystem, enabling it to address physics-based machine learning, optimization, and inverse problems. XLB has been successfully scaled to handle simulations with billions of cells, achieving giga-scale lattice updates per second. XLB is released under the permissive Apache-2.0 license and is available on GitHub at https://github.com/Autodesk/XLB. Program Title: XLB CPC Library link to program files: https://doi.org/10.17632/n25h65vw59.1 Developer's repository link: https://github.com/Autodesk/XLB Licensing provisions: Apache-2.0 Programming language: Python Nature of problem: XLB is a scalable Lattice Boltzmann simulation library based on JAX. It is designed to be accessible, extensible, and high performance. The library can be readily augmented with novel boundary conditions, collision models, or simulation capabilities. Moreover, XLB is differentiable, offering the unique advantage of integration with JAX's extensive machine learning ecosystem, and tackling physics-based machine learning, optimization, and inverse problems. Solution method: XLB utilizes the JAX library for just-in-time (JIT) compilation and distributed computing, enabling it to scale effectively across distributed multi-GPU/multi-TPU systems. XLB's programming model combines an object-oriented interface for easy extensibility. The library can be readily augmented with novel boundary conditions, collision models, or other simulation capabilities. [ABSTRACT FROM AUTHOR]

RHEED_DIFF_2D is an open-source software capable of performing qualitative numerical simulations of changes to the intensity of RHEED oscillations in the function of the layer deposited, employing various models of scattering crystal potential for heteroepitaxial structures of bcc (001) films nucleated on a fcc (001) surface, including the possible existence of various diffuse scattering models through the layer parallel to the surface. This model can be used for interpreting data weakly dependent on the azimuth of the incident electron beam direction and for 2D monolayer-by-monolayer growth modes. The program does not use any external data sources. Program title: RHEED_DIFF_2D CPC Library link to program files: https://doi.org/10.17632/83ht2kc8xt.1 Licensing provisions: GNU General Public License 3 Programming language used: C++ Journal reference of previous version: Computer Physics Communications 185 (2014) 3001–3009 Does the new version supersede the previous version?: No. It is a supplement to the previous version. [ABSTRACT FROM AUTHOR]

This paper proposes a new software implementation for the rapid development and deployment of computer simulations of reflection high-energy electron diffraction (RHEED) intensity oscillations during the growth of thin epitaxial films. The calculations are based on the kinematical approximation. In this model the RHEED intensity oscillations from growing layers is considered as an effect of changes of surface roughness. Program Title: growth22 CPC Library link to program files: https://doi.org/10.17632/zkpwhrrfng.1 Licensing provisions: GNU General Public License 3 Programming language: C++ Journal reference of previous version: Computer Physics Communications 170 (2005) 265–286 Does the new version supersede the previous version?: Yes Reasons for the new version: The previous version of the program was written in Object Pascal [1]. This programming language is no longer popular in the scientific community. Responding to user's feedback, the current distribution of the software has been implemented in C++. This version implements improvements for ergonomics, computational performances and code readability. In addition, the output generated by the growth22 (saved in the coverageProfiles.dat file) can be used directly by program [2]. Summary of revisions: The growth22 is written in standard C++, therefore one can use the GNU g++ compiler or any ANSI C++ compliant compiler to build the executable program. The code is distributed in the form of one growth22.cpp source file. A Table 1 lists the retired Object Pascal procedures and shows replacements by C++ functions. The present version of the program enables the user to easily manage the input data. Input data are loaded from a manually prepared inputData.dat text file where the user provides information on the parameters of the simulation, i.e. the growth model, the number of growing layers, the upper limit of growth time interval, the values of the phenomenological parameters that measure the net rate of transfer from one layer to the next, and the values of the growth rate for each layer. The user can edit this file as per the requirement of the problem domain. Details of the parameters in the sample inputData.dat file are described in the readme.pdf file provided with the software. The nature of problem: Generally, in order to explain the experimental data two theories—kinematical and dynamical—associated with high-energy electron diffraction in epitaxial layers are used. The dynamical theories take into account multiple scattering within the crystal and are used whenever diffraction from a perfect epitaxial layers is considered [3]. On the other hand, the kinematical theory treats the scattering from each plane in the crystal sample as being independent of others [1,4]. In the kinematical theory of RHEED it is assumed that electron wave scatters from atom in the surface layer only once and then interferes with waves scattered on other atoms. This model can be used to describe diffraction effects for small angles of incidence, when the electron beam does not penetrate the deeper layers of the crystal. In the kinematical model the intensity oscillations of specularly reflected electron beam from growing layer is considered as an effect of changes of surface roughness. In this paper we present a time-efficient computer model for kinematical calculations of RHEED intensity oscillations during the growth of thin epitaxial films. The method of solution: According to kinematical approximation the diffracted-beam intensity is given by [4]: (1) I ( S ‖ → = 0 , S z) = | ∑ n = 0 ∞ (θ n − θ n + 1) e x p (− i S z n d) | 2 where S → is the scattering vector, d is the interlayer spacing and n is an integer. θ n − θ n + 1 is the exposed layer coverage which is treated as the atomic distribution function gives the position of surface atoms. For S z = π / d Eq. (1) simplifies to: (2) I = | ∑ n = 0 ∞ (θ n − θ n + 1) cos ⁡ (n π) | 2. Variation of the coverage ratio θ n of the n th surface layer is modelled by a set of non-linear differential equations [1,4]. The Cash-Karp fifth-order Runge-Kutta method [5] with error estimation and adaptive stepsize control was used for solving initial value problem for non-linear differential equations [6]. The absolute and relative error tolerances and number of integration intervals are fully user-configurable. The proposed code approaches the modelling of RHEED from two different points of view, i.e. the data and functional perspectives. The data perspective (i.e. calculating the layer coverages as a function of their growth time) is used to characterize the data sources of the program whereas the functional perspective includes services provided by the program. It should be noted that the perspective of the data can be expressed by a completely different model introduced by the user. The coverage evolution curves for the diffusive growth model, corresponding kinematical diffracted intensity and total roughness of the surface are shown in Fig. 1. During the numerical calculations showed in Fig. 1 it was assumed that the number of growing layers (numLayers) equals 15, the upper limit of growth time interval (tMax) equals 12, the growth rate equals 1 MLs−1 (for all layers), and the values of the parameter k n (that determines the amount of interlayer diffusion) changing according to the linear relation: (3) k n = k 1 − (k 1 − k N) n − 1 N − 1 , where N is the total number of growing layers. [1] A. Daniluk, Comput. Phys. Commun. 170 (2005) 265. [2] A. Daniluk, I. Żurawek, R. Zdyb, Comput. Phys. Commun. 261 (2021) 107692. [3] T. Kawamura, P.A. Maksym, Surf. Sci. 630 (2014) 125. [4] P.I. Cohen, G.S. Petrich, P.R. Pukite, G.J. Whaley, A.S. Arrott, Surf. Sci. 216 (1989) 222. [5] J. R. Cash, A. H. Karp, ACM Trans. Math. Softw. 16 (1990) 201. [6] W.H. Press, B.P. Flannery, S.A. Teukolsky, W.T. Vetterling, Numerical Recipes. The Art of Scientific Computing, third ed., Cambridge University Press, 2007; See also: Numerical Recipes in C, second ed., Cambridge University Press, 1992. [ABSTRACT FROM AUTHOR]

In recent years graphical processing units (GPUs) have become a powerful tool in scientific computing. Their potential to speed up highly parallel applications brings the power of high performance computing to a wider range of users. However, programming these devices and integrating their use in existing applications is still a challenging task. In this paper we examined the potential of GPUs for two different applications. The first application, created at Paul Scherrer Institut (PSI), is used for parameter fitting during data analysis of μ SR (muon spin rotation, relaxation and resonance) experiments. The second application, developed at ETH, is used for PET (Positron Emission Tomography) image reconstruction and analysis. Applications currently in use were examined to identify parts of the algorithms in need of optimization. Efficient GPU kernels were created in order to allow applications to use a GPU, to speed up the previously identified parts. Benchmarking tests were performed in order to measure the achieved speedup. During this work, we focused on single GPU systems to show that real time data analysis of these problems can be achieved without the need for large computing clusters. The results show that the currently used application for parameter fitting, which uses OpenMP to parallelize calculations over multiple CPU cores, can be accelerated around 40 times through the use of a GPU. The speedup may vary depending on the size and complexity of the problem. For PET image analysis, the obtained speedups of the GPU version were more than × 40 larger compared to a single core CPU implementation. The achieved results show that it is possible to improve the execution time by orders of magnitude. [ABSTRACT FROM AUTHOR]

This paper introduces Eilmer, a general-purpose open-source compressible flow solver developed at the University of Queensland, designed to support research calculations in hypersonics and high-speed aerothermodynamics. Eilmer has a broad userbase in several university research groups and a wide range of capabilities, which are documented on the project's website, in the accompanying reference manuals, and in an extensive catalogue of example simulations. The first part of this paper describes the formulation of the code: the equations, physical models, and numerical methods that are used in a basic fluid dynamics simulation, as well as a handful of optional multi-physics models that are commonly added on to do calculations of hypersonic flow. The second section describes the processes used to develop and maintain the code, documenting our adherence to good programming practice and endorsing certain techniques that seem to be particularly helpful for scientific codes. The final section describes a half-dozen example simulations that span the range of Eilmer's capabilities, each consisting of some sample results and a short explanation of the problem being solved, which together will hopefully assist new users in beginning to use Eilmer in their own research projects. Program Title: Eilmer CPC Library link to program files: https://doi.org/10.17632/gy2ds2fyxm.1 Developer's repository link: https://github.com/gdtk-uq/gdtk Code Ocean capsule: https://codeocean.com/capsule/7226427 Licensing provisions: GPLv3 Programming language: D, Lua Supplementary material: https://gdtk.uqcloud.net Nature of problem: Eilmer solves the compressible Navier-Stokes equations with a particular emphasis on flows at hypersonic speeds. The code includes modelling for high-temperature gas effects such as chemical and vibrational nonequilibrium. Eilmer can be used for the simulation for unsteady and steady flows. Solution method: The code is implemented in D [1] and built on a finite-volume formulation that is capable of solving the Navier-Stokes equations in 2D and 3D computational domains, discretised with structured or unstructured grids. Grids may be generated using a built-in parametric scripting tool or imported from commercial gridding software. The inviscid fluxes are computed using the reconstruction-evolution approach. In structured-grid mode, reconstruction stencils up to fourth-order spatial accuracy are available. In unstructured-grid mode, least-squares reconstruction provides second-order spatial accuracy. A variety of flux calculators are available in the code. Viscous fluxes are computed with compact stencils with second-order spatial accuracy. For unsteady flows, explicit time-stepping with low-order RK-family schemes are available, along with a point-implicit Backward-Euler update scheme for stiff systems of equations. For steady flows, convergence can be greatly accelerated using a Jacobian-free Newton-Krylov update scheme, which seeks a global minimum in the residuals using a series of large pseudo-timesteps. Domain decomposition is used for parallel execution using both shared memory and distributed memory programming techniques. Additional comments including restrictions and unusual features: Eilmer provides a programmable interface for pre-processing, post-processing and user run-time customisations. The programmable interface is enabled using a built-in embedded interpreter for the Lua programming language [2]. Run-time customisations include used-defined boundary conditions, source terms and grid motion. [1] D Programming Language web page: https://dlang.org/. [2] Lua Programming Language web page: https://www.lua.org/. [ABSTRACT FROM AUTHOR]

A computational fluid dynamics (CFD) simulation framework for fluid-flow prediction is developed on the Tensor Processing Unit (TPU) platform. The TPU architecture is featured with accelerated dense matrix multiplication, large high bandwidth memory, and a fast inter-chip interconnect, making it attractive for high-performance scientific computing. The CFD framework solves the variable-density Navier-Stokes equation using a low-Mach approximation, and the governing equations are discretized by a finite-difference method on a collocated structured mesh. It uses the graph-based TensorFlow as the programming paradigm. The accuracy and performance of this framework is studied both numerically and analytically, specifically focusing on effects of TPU-native single precision floating point arithmetic. The algorithm and implementation are validated with canonical 2D and 3D Taylor-Green vortex simulations. To demonstrate the capability for simulating turbulent flows, simulations are conducted for two configurations, namely decaying homogeneous isotropic turbulence and a turbulent planar jet. Both simulations show good statistical agreement with reference solutions. The performance analysis shows a linear weak scaling and a superlinear strong scaling up to a full TPU v3 pod with 2048 cores. [ABSTRACT FROM AUTHOR]

A practical computing algorithm has been developed for calculating the reflection high-energy electron diffraction (RHEED) from the molecular beam epitaxy growing surface. The calculations are based on the use of the dynamical diffraction theory in which the electrons are taken to be diffracted by a potential, which is periodic in the dimension perpendicular to the surface. The computer program presented in this paper enables calculations for three basic types of diffuse potential for crystalline heteroepitaxial structures, including the possible existence of various diffuse scattering models through the layer parallel to the surface. Program Summary Program title: RHEED_DIFF Catalogue identifier: AETW_v1_0 Program summary URL: http://cpc.cs.qub.ac.uk/summaries/AETW_v1_0.html Program obtainable from: CPC Program Library, Queen’s University, Belfast, N. Ireland Licensing provisions: Standard CPC licence, http://cpc.cs.qub.ac.uk/licence/licence.html No. of lines in distributed program, including test data, etc.: 68483 No. of bytes in distributed program, including test data, etc.: 480831 Distribution format: tar.gz Programming language: C + + . Computer: Intel i7-based PC. Operating system: Windows, Linux. RAM: The presented algorithm belongs to the linear memory class of the computational complexity O ( n ) . Word size: 64 bits Classification: 4.6, 6.2, 7.2. Nature of problem: RHEED rocking curves (the specular beam intensities versus the glancing angle) recorded from heteroepitaxial layers are used for the non-destructive evaluation of epilayer thickness and composition with a high degree of accuracy. Rocking curves from such heterostructures are often very complex because the thickness fringes from every layer beat together. Simulations based on the dynamical diffraction theory are generally used to interpret the rocking curves of such structures from which very small changes in thickness and composition can be obtained. Rocking curves are also used to determine the level of strain and its relaxation mechanism in a lattice-mismatched system. Solution method: RHEED intensities are calculated within the framework of the general matrix formulation described in Ref. [1] under the one-beam condition [2,3] . The dynamical diffraction calculations presented in this paper utilize the systematic reflection case in RHEED, in which the atomic potential in the planes parallel to the surface are projected onto the surface normal, so that the results are insensitive to the atomic arrangement in the layers parallel to the surface. In this paper, an approach in which oscillating changes in the intensity appear as a combined effect of changes in the refraction conditions and changes in diffuse scattering for different models of the scattering potential is implemented. Running time: Numerically, the problem of calculating the changes in RHEED oscillation intensity from growing layers is an NP problem. The time computational complexity of the presented algorithm depends on the number of layers for both the substrate and the growing layers included in the calculations. The time-computational complexity of the presented solution is O ( n 2 ) , where n is the total number of layers used in the calculations. References: [1] A. Daniluk, Comput. Phys. Commun. 166 (2005) 123. [2] L.M. Peng, M.J. Whelan, Surf. Sci. Lett. 238 (1990) L446. [3] Y. Horio, A. Ichimiya, Japan. J. Appl. Phys. 33 (1994) L377. [ABSTRACT FROM AUTHOR]

The use of flux limiters is widespread within the scientific computing community to capture shock discontinuities and are of paramount importance for the temporal integration of high-speed aerodynamics, multiphase flows and hyperbolic equations in general. Meanwhile, the breakthrough of new computing architectures and the hybridization of supercomputer systems pose a huge portability challenge, particularly for legacy codes, since the computing subroutines that form the algorithms, the so-called kernels, must be adapted to various complex parallel programming paradigms. From this perspective, the development of innovative implementations relying on a minimalist set of kernels simplifies the deployment of scientific computing software on state-of-the-art supercomputers, while it requires the reformulation of algorithms, such as the aforementioned flux limiters. Equipped with basic algebraic topology and graph theory underlying the classical mesh concept, a new flux limiter formulation is presented based on the adoption of algebraic data structures and kernels. As a result, traditional flux limiters are cast into a stream of only two types of computing kernels: sparse matrix-vector multiplication and generalized pointwise binary operators. The newly proposed formulation eases the deployment of such a numerical technique in massively parallel, potentially hybrid, computing systems and is demonstrated for a canonical advection problem. [ABSTRACT FROM AUTHOR]

This paper presents a computer program, which facilitates the calculation of changes to the intensity of RHEED oscillations from the heteroepitaxial structures of (0001)GaN films nucleated on a Si surface. The calculations are based on the use of a dynamical diffraction theory and different models of scattering crystal potential. New version program summary Title of program: RHEED_DIFF_W Program Files doi: 10.17632/hn3pt6ytky.1 Licensing provisions: GNU General Public License 3 Programming language used: C++ Journal reference of previous version: Computer Physics Communications 185 (2014) 3001–3009 Does the new version supersede the previous version?: No. It does supersede version AETW_v1_0. Reasons for the new version: Responding to the users’ feedback we presented a practical procedure of construction of simulation program, which facilitates the calculation of changes to the intensity of RHEED oscillations in the function of the glancing angle of incidence of the electron beam, employing various models of scattering crystal potential for heteroepitaxial structures of hexagonal (0001)GaN films nucleated on a Si surface. GaN has a wurtzite crystal structure with alternating planes of Ga and N atoms and lattice constants a = 3.189 Å and c = 5.185 Å [1]. Choosing the top surface to be z = 0 , the N atoms are located at z = n d where d = 5.185 / 2 Å and n is an integer, and Ga atoms are at z = n d − u , where u = 3 c / 8 (as shown in Fig. 1 ). The nature of physical problem: Since the discovery of molecular beam epitaxy, which allows formation of layers (crystals) with a controlled composition of subsequent monolayers, the research attempts of many laboratories have been focused on studying the dynamics of formation of both monolayers and specific heterojunction structures, that is the structures with monocrystalline interfaces created by materials of various chemical compositions (different physical and chemical properties). The fundamental scientific problem in such research is to specify both interface type and growth mechanism for subsequent layers. Researchers and technologists manufacturing two-dimensional nanoscale systems frequently use RHEED rocking curves (the specular beam intensities versus the glancing angle) to control growth of films at the atomistic level of accuracy. The structural properties of the GaN/Si interface have been meticulously studied by many [3–5], and crucial problem of the direct growth of GaN on Si is the formation of amorphous silicon nitride (SiN) at the GaN/Si interface. Typical running time: The typical running time is machine and user-parameters dependent, i.e. the number of layers for both the substrate and the growing layers included in the calculations. The method of solution: RHEED intensities are calculated within the general framework described in Ref. [2] with the model of the scattering potential for heterostructures: (1) U C o m b i n e d ( z ) = ∑ n U s u b s t r a t e ( z ) + ∑ n ( U l a y e r ( z ) + U a d d l a y e r ( z ) ) , where the U a d d l a y e r represents the diffuse scattering on the topmost layer. Summary of revisions: The presented version of computer program implements original algorithms of self-consistent calculations for scattering potentials GaN(0001)/Si(111) and solving a time-independent Schrödinger equation for high-energy electrons ( Fig. 2 ). Fig. 1 A plot of the real part of the one-dimensional potential of GaN(0001) for 12 layers thickness. Fig. 2 The specular beam intensities versus the glancing angle for the GaN(0001) corresponds to the data shown in Fig. 1 . Electron energy equals 10 keV. Fig. 3 The one-dimensional (real part) potential of GaN(0001)/Si(111) for 10 layer thickness of Si and one of GaN corresponds to the data shown in Fig. 4 . Fig. 4 A set of rocking curves calculated during the growth of a single GaN layer on the Si(111) substrate for two different values θ G a N of the coverage of the topmost layer. For the calculations the MODEL3 [2] of the scattering potential was used. During the numerical calculations of the changes in the intensity of the specular beam in the function of the glancing angle, it was assumed that the electron energy equals 10 keV, the glancing angle was increased from 0.5° to 7°, and the value of α parameter = 0.1 and β = 0.5 for the MODEL3 of the scattering potential [2]. Figs. 3 – 4 present the situation where on the crystalline Si(111) substrate a single GaN layer was placed, where the GaN growth on Si is having a thin amorphous layer. It can be observed that for a very thin layer of GaN, a clear minimal amplitude of RHEED oscillation occurs with the coverage of the GaN layer equal to about 0.5. References [1] A. Ichimiya, P.I. Cohen, Reflection High-Energy Electron Diffraction, Cambridge University Press, Cambridge, 2004. [2] A. Daniluk, Comput. Phys. Commun. 185 (2014) 3001. [3] M. Tamura, M. López-López, Superficies y Vacío 13 (2001) 80. [4] J. Liang, S.-K. Hong, N. Kouklin, R. Beresford, and J.M. Xu, Appl. Phys. Lett. 83(9) (2003) 1752. [5] M. Sobanska, K. Klosek, Z.R. Zytkiewicz, J. Borysiuk, B.S. Witkowski, E. Lusakowska, A. Reszka, and R. Jakiela, Cryst. Res. Technol. 47(3) (2012) 307. [ABSTRACT FROM AUTHOR]

Abstract: The method of discontinuous finite element discrete ordinates which involves inverting an operator by iteratively sweeping across a mesh from multiple directions is commonly used to solve the time-dependent particle transport equation. Graphics Processing Unit (GPU) provides great faculty in solving scientific applications. The particle transport with unstructured grid bringing forward several challenges while implemented on GPU. This paper presents an efficient implementation of particle transport with unstructured grid under 2D cylindrical Lagrange coordinates system on a fine-grained data level parallelism GPU platform from three aspects. The first one is determining the sweep order of elements from different angular directions. The second one is mapping the sweep calculation onto the GPU thread execution model. The last one is efficiently using the on-chip memory to improve performance. As to the authorsʼ knowledge, this is the first implementation of a general purpose particle transport simulation with unstructured grid on GPU. Experimental results show that the performance speedup of NVIDIA M2050 GPU with double precision floating operations ranges from 11.03 to 17.96 compared with the serial implementation on Intel Xeon X5355 and Core Q6600. [Copyright &y& Elsevier]

Abstract: This paper presents the problem of Molecular Beam Epitaxy and Reflection High-Energy Electron Diffraction with the help of a unified, modern MDA approach. Model-Driven Architecture (MDA) constitutes a modern and unusually efficient method of improving the process of generating software. It was created at the beginning of the twenty-first century by the Object Management Group as an element of Model-Driven Development, a highly promoted trend in software engineering. In MDA a viewpoint on a system is a technique for abstraction using a selected set of architectural concepts and structuring rules, in order to focus on particular concerns within a system. In MDA, system design begins with defining the problem domain. Next, at a highly abstract level—independent of the system and programming platform—a Platform-Independent Model (PIM) is constructed as well as a general system specification. This specification is created with the help of Unified Modeling Language. The real implementation of the system is performed through the transformation of PIM to Platform-Specific Model (PSM). The essence of Model-Driven Architecture is the replacement of the twentieth century approach to programming, calling that “everything is an object”, to the modern—“everything is a model”. [Copyright &y& Elsevier]

In atomic structure and collision theory, the efficient spin-angular integration is known to be crucial and often decides, how accurate the properties and behavior of atoms can be predicted numerically. Various methods have been developed in the past to keep the computation (and implementation) of the spin-angular integration feasible for complex shell structures, including open d - and f -shell elements. To support such computations, we here provide a new implementation of the angular coefficients for jj -coupled and symmetry-adapted configuration states that is entirely built upon the quasi-spin formalism. The module SpinAngular is based on Julia, a new programming language for scientific computing, and supports a simple access to all (completely) reduced tensors, coefficients of fractional parentage for subshells with j ≤ 9 / 2 as well as the re-coupling coefficients from this formalism. Moreover, this module has been worked out for multiple purposes, including 1) the accurate calculation of atomic properties, 2) further studies on spin-angular integration theory, 3) the development of new or existing computer programs as well as 4) the manipulation of reduced matrix elements from this theory. The present implementation will therefore help advance the algebraic evaluation of many-electron (transition) amplitudes and to apply the theory to newly emerging research areas. Program title: SpinAngular CPC Library link to program files: https://doi.org/10.17632/jjpff3pysn.1 Code Ocean capsule: https://doi.org/10.24433/CO.3772334.v1 Licensing provisions: MIT License Programming language: Julia Nature of problem: For symmetric one- and two-particle operators, the (many-electron) matrix elements between symmetry-adapted configuration states can always be written as a sum of angular coefficient x interaction strength. While the interaction strengths describe the physical interaction and just depends on the (one-electron) orbitals of the active shells, the angular coefficients arise from the (many-electron) spin-angular integrals and are determined geometrically by the occupation and coupling of all electrons. These angular coefficients need to be calculated efficiently and often require special care. Solution method: The spin-angular approach [1] is implemented for (tensorial coupled) one-particle operators of arbitrary rank, scalar two-particle operators as well as for the manipulation of reduced matrix elements from this theory. [1] G. Gaigalas, Z. Rudzikas, C. Froese Fischer, J. Phys. B, At. Mol. Opt. Phys. 30 (1997) 3747. [ABSTRACT FROM AUTHOR]

Software to analyse very large sets of experimental data often relies on a pipeline of irregular computational tasks with decisions to remove irrelevant data from further processing. A user-centred framework was designed and deployed, HEP-Frame, which aids domain experts to develop applications for scientific data analyses and to monitor and control their efficient execution. The key feature of HEP-Frame is the performance portability of the code across different heterogeneous platforms, due to a novel adaptive multi-layer scheduler, seamlessly integrated into the tool, an approach not available in competing frameworks. The multi-layer scheduler transparently allocates parallel data/tasks across the available heterogeneous resources, dynamically balances threads among data input and computational tasks, adaptively reorders in run-time the parallel execution of the pipeline stages for each data stream, respecting data dependencies, and efficiently manages the execution of library functions in accelerators. Each layer implements a specific scheduling strategy: one balances the execution of the computational stages of the pipeline, distributing the execution of the stages of the same or different dataset elements among the available computing threads; another controls the order of the pipeline stages execution, so that most data is filtered out earlier and later stages execute the computationally heavy tasks; yet another adaptively balances the automatically created threads among data input and the computational tasks, taking into account the requirements of each application. Simulated data analyses from sensors in the ATLAS Experiment at CERN evaluated the scheduler efficiency, on dual multicore Xeon servers with and without accelerators, and on servers with the many-core Intel KNL. Experimental results show significant improved performance of these data analyses due to HEP-Frame features and the codes scaled well on multiple servers. Results also show the improved HEP-Frame scheduler performance over the key competitor, the HEFT list scheduler. The best overall performance improvement over a real fine tuned sequential data analysis was impressive in both homogeneous and heterogeneous multicore servers and in many-core servers: 81x faster in the homogeneous 24+24 core Skylake server, 86x faster in the heterogeneous 12+12 core Ivy Bridge server with the Kepler GPU, and 252x faster in the 64-core KNL server. Program Title: HEP-Frame CPC Library link to program files: https://doi.org/10.17632/m2jwxshtfz.1 Licencing provisions: GPLv3 Programming language: C++. Supplementary material: The current HEP-Frame public release available at https://bitbucket.org/ampereira/hep-frame/wiki/Home. Nature of problem: Scientific data analysis applications are often developed to process large amounts of data obtained through experimental measurements or Monte Carlo simulations, aiming to identify patterns in the data or to test and/or validate theories. These large inputs are usually processed by a pipeline of computational tasks that may filter out irrelevant data (a task and its filter is addressed as a proposition in this communication), preventing it from being processed by subsequent tasks in the pipeline. This data filtering, coupled with the fact that propositions may have different computational intensities, contribute to the irregularity of the pipeline execution. This can lead to scientific data analyses I/O-, memory-, or compute-bound performance limitations, depending on the implemented algorithms and input data. To allow scientists to process more data with more accurate results their code and data structures should be optimized for the computing resources they can access. Since the main goal of most scientists is to obtain results relevant to their scientific fields, often within strict deadlines, optimizing the performance of their applications is very time consuming and is usually overlooked. Scientists require a software framework to aid the design and development of efficient applications and to control their parallel execution on distinct computing platforms. Solution method: This work proposes HEP-Frame, a framework to aid the development and efficient execution of pipelined scientific analysis applications on homogeneous and heterogeneous servers. HEP-Frame is a user-centred framework to aid scientists to develop applications to analyse data from a large number of dataset elements, with a flexible pipeline of propositions. It not only stresses the interface to domain experts so that code is more robust and is developed faster, but it also aims high-performance portability across different types of parallel computing platforms and desirable sustainability features. This framework aims to provide efficient parallel code execution without requiring user expertise in parallel computing. Frameworks to aid the design and deployment of scientific code usually fall into two categories: (i) resource-centred, closer to the computing platforms, where execution efficiency and performance portability are the main goals, but forces developers to adapt their code to strict framework constraints; (ii) user-centred, which stresses the interface to domain experts to improve their code development speed and robustness, aiming to provide desirable sustainability features but disregarding the execution performance. There are also a set of frameworks that merge these two categories (Liu et al., 2015 [1] ; Deelman et al., 2015 [2]) for scientific computing. While they do not have steep learning curves, concessions have to be made to their ease of use to allow for their broader scope of targeted applications. HEP-Frame attempts to merge this gap, placing itself between a fully user- or resource-centred framework, so that users develop code quickly and do not have to worry about the computational efficiency of the code It handles (i) by ensuring efficient execution of applications according to their computational requirements and the available resources on the server through a multi-layer scheduler, while (ii) is addressed by automatically generating code skeletons and transparently managing the data structure and automating repetitive tasks. Additional comments: An early stage proof-of-concept was published in a conference proceedings (Pereira et al., 2015). However, the HEP-Frame version presented in this communication only shares a very small portion of the code related to the skeleton generation (less than 5% of the overall code), while the rest of the user interface, multi-layer scheduler, and parallelization strategies were completely redesigned and re-implemented. • A multi-layer scheduler to efficiently execute irregular pipelined data analyses. • Run-time balance #threads for simultaneous data read/setup and pipeline processing. • Scheduling data analyses on single or multiple homogeneous/heterogeneous servers. • Adaptive run-time reordering of parallel pipeline stages on multiple dataset values. • Impressive speedups (up to 252x) for compute/memory-bound real world applications. [ABSTRACT FROM AUTHOR]

We report a simulation program which facilitates the calculation of changes in the intensity of specular reflection of electron beams in RHEED experiments for thin epitaxial films deposited on crystalline surfaces. It has been shown that the amplitude of the RHEED intensity oscillations greatly depends on the glancing angle of the incident electron beam, the coverages of the growing layers and the model of the scattering potential. The usefulness of the program has been tested on a well-known system of Ag grown on a Si(111) surface. The obtained experimental and computational results correspond closely. The presented algorithm, together with properly modified input data, can be applied to other systems of crystalline ultrathin layer and substrate. It also enables the implementation and tests of different combinations of the scattering potentials of the crystal, and can be applied to interpret experimental RHEED rocking curves. Program title: RHEED_DIFFv2 CPC Library link to program files: https://doi.org/10.17632/7c6y233rys.1 Code Ocean capsule: https://codeocean.com/capsule/0078674 Licensing provisions: GNU General Public License 3 Programming language: C++ Journal reference of previous version: Computer Physics Communications 185 (2014) 3001–3009 Does the new version supersede the previous version?: Yes Reasons for the new version: Responding to user's feedback we improved functionality of the program. Moreover, we added new capabilities which make the input data process easier and more efficient than the previous one. Nature of problem: The measurement of incident-angle dependence of RHEED intensity oscillations (RHEED rocking curve) during the growth of thin layers prepared by molecular beam epitaxy is a popular technique for quantitative and qualitative investigations of epitaxial structure perfection. Rocking curves recorded from heteroepitaxial layers are used for the non-destructive characterization of atoms near the surface [1,2]. Rocking curves are also used to determine the level of strain and its relaxation mechanism in lattice-mismatched systems. In most cases the interpretation of experimental results is based on the use of dynamical diffraction approaches. Such approaches are known to be useful in qualitative and quantitative analyses of RHEED experimental data, and especially in the interpretation of changes in intensity of reflected specular electron beam. Solution method: RHEED intensities are calculated within the general framework described in Ref. [3] with the model of the scattering potential for heterostructures: (1) U (θ , z) = ∑ i ∑ n U n substrate (θ (n) = 1 , z i) + ∑ i ∑ n (U n layer (θ (n) , z i) + U add layer (θ (n) , z i)) , where U n substrate means the potential of the full-filled substrate layers, U n layer means the potential of growing monolayer, U add layer component is responsible for diffuse scattering, and θ (n) is the value of n th layer coverage in the vicinity of a growth front as a function of deposited layers [4]. The description of atom deposition at growing surfaces was obtained by solving the set of nonlinear differential equations developed for the distributed growth model [4]. (2) d θ n (t) dt = 1 τ n [ θ n − 1 − θ n + α n (θ n − θ n + 1) − α n − 1 (θ n − 1 − θ n) ] , where (3) α n = A n d n (θ n) d n (θ n) + d n + 1 (θ n + 1) , and (4) d n (θ n) = θ (1 − θ n) . In these equations, A n is the parameter that measures the net rate transfer of atoms from one layer to the next, and τ n is the deposition time of n th monolayer. Detailed explanations for Eqs. (2-4) are presented in Ref. [4]. The current distribution of the RHEED_DIFFv2 program includes an example input file CoverageProfiles.dat. It is one of the program output files [4] (but it can also be prepared independently of the program [4]). The CoverageProfiles.dat file contains coverage profiles for numerical solutions of Eqs. (2-4) with parameters 1 ∕ τ n<=2 = 1. 0 , 1 ∕ τ n>=3 = 0. 8 , and A 1 = 0. 999 , A 2 = 0. 6 , A 3 = 0. 9 , A 4 = 0. 9 , A 5 = 0. 8 , A 6 = 0. 7 , A n>=7 = 0. 6. The values of these parameters correspond to the Ag/Si(111) growth model proposed by Zhang and co-workers [5]. Summary of revisions: In the previous version of the program for calculating the amplitude of the reflected electron beam from growing layers, only the growth mode for the topmost surface monolayer has been taken into account [3]. In the current version it is possible to include in the calculations all growing monolayers. For this purpose, we have implemented an original algorithm for dynamical calculations of changes in the RHEED rocking curve for selected growth model of thin epitaxial films [4]. In the current version of the program, the scattering potentials of successive growing monolayers are modified by the coverage values of the monolayer near the growth front. Therefore, the program needs CoverageProfiles.dat file [4] with input data to work properly. This file stores the values of layer coverages in the vicinity of a growth front as a function of deposited layers. The data in this file are saved in two columns — the first column stores the ordinal number of deposited layers, while the second contains the values of the growth front for each of these layers. Fig. 1 (a) shows experimentally measured rocking curves for Ag layers (2.83 nm-thick) for the azimuth of incidence [11-2] and [11-2]+7°. For the [11-2] direction, and turned away by 7° from [11-2], the three dimensional crystal lattice of the Si(111) substrates and Ag(111) growing layers can be regarded as a one-dimensional array of lattice planes parallel to the surface. This azimuthal direction corresponds to the one-beam condition [3]. Figs. 1 (b–c) show dynamically calculated one-beam rocking curves for Ag(111)/Si(111) layers. In the numerical calculations of the rocking curves of specular beam intensity for 2.83 nm-thick Ag layers, we used the following parameters: the electron energy of 18.8 keV, the glancing angle in the range from 0.2° to 3.5°, and the values α = 0. 2 , β = 0. 0 and 0.5 for the model of the scattering potential described by Eq. (1). Assuming that results of one-beam calculations reproduce actual experimental situations, we can conclude that for a fixed, real surface, the measured and calculated positions of Bragg reflections tally very well. The difference between them does not exceed 0.2°, which remains within the limits of experimental error. Acknowledgments This work has been in part supported by the National Science Centre under Grant No. 2016/21/B/ST3/01294. References Y. Horio, J. Yuhara, and Y. Takakuwa, Jap. J. Appl. Phys. 58 (2019) SIIA14 T. Kawamura, K. Fukutani, Surf. Sci. 688 (2019) 7-13 A. Daniluk, Comput. Phys. Commun. 185 (2014) 3001-3009. A. Daniluk, Comput. Phys. Commun. 170 (2005) 265–286. Z. H. Zhang, S. Hasegawa, and S. Ino, Phys. Rev. B 55 (1997) 9983. [ABSTRACT FROM AUTHOR]

Nonequispaced discrete Fourier transformation (NDFT) is widely applied in all aspects of computational science and engineering. The computational efficiency and accuracy of NDFT has always been a critical issue in hindering its comprehensive applications both in intensive and in extensive aspects of scientific computing. In our previous work Yang et al. (2018), a CUNFFT method was proposed and it shown outstanding performance in handling NDFT at intermediate scale based on CUDA (Compute Unified Device Architecture) technology. In the current work, we further improved the computational efficiency of the CUNTTF method using an efficient MPI-CUDA hybrid parallelization (HP) scheme of NFFT to achieve a cutting-edge treatment of NDFT at super extended scale. Within this HP-NFFT method, the spatial domain of NDFT is decomposed into several parts according to the accumulative feature of NDFT and the detailed number of CPU and GPU nodes. These decomposed NDFT subcells are independently calculated on different CPU nodes using a MPI process-level parallelization mode, and on different GPU nodes using a CUDA thread-level parallelization mode and CUNFFT algorithm. A massive benchmarking of the HP-NFFT method indicates that this method exhibits a dramatic improvement in computational efficiency for handling NDFT at super extended scale without loss of computational precision. Furthermore, the HP-NFFT method is validated via the calculation of Madelung constant of fluorite crystal structure, and thereafter verified that this method is robust for the calculation of electrostatic interactions between charged ions in molecular dynamics simulation systems. Program title: HP-NFFT CPC Library link to program files: http://dx.doi.org/10.17632/ys2y92jkwy.1 Licensing provisions: GNU General Public License 3 Programming language: MPI, C, and CUDA C Supplementary material: The program is designed for effective computation of large-scale nonequispaced discrete Fourier transformation (NDFT), which runs on particular computers equipped with NVIDIA GPUs. It has been tested on (a) one single computer node with Intel(R) Core(TM) i7-3770 @ 3.40 GHz (CPU) and GTX 980 Ti (GPU), and (b) MPI parallel computer nodes with the same configurations. Nature of problem: For NDFT, the computation is extremely time-consuming in many domains of computational physics due to the failure to utilize FFT directly, which often affects the computational efficiency of whole system seriously. Although the parallel method CUNFFT based on GPU has achieved a qualitative leap compared with previous methods in NDFT computation, the computation capability is limited to the throughput capacity of GPU when the size of NDFT system is big enough. Solution method: We constructed a hybrid parallel architecture, in which CPU and GPU are combined to accelerate the NDFT computation effectively. Firstly, the NDFT system is divided into several subcells via domain-decomposition method. Then MPI (Message Passing Interface) is used to implement the CPU-parallel computation with each computer node corresponding to a particular subcell, and furthermore each subcell in one computer node will be executed in GPU in parallel efficiently. In this hybrid parallel method, the most critical technical problem is how to parallelize a CUNFFT in the parallel strategy, which is conquered effectively by deep-seated research of basic principles and some algorithm skills. Restrictions: The HP-NFFT is mainly oriented to, and has shown significant computational efficiency to deal with large-scale NDFT computations. However, for a small NDFT system containing less than 10^6 particles, the mode of multiple computer nodes has no apparent efficiency advantage or even lower efficiency compared with the mode of a single computer node because of the serious network delay. References: [1] S.-C. Yang, H.-J. Qian, Z.-Y. Lu, Appl. Comput. Harmon. Anal. 44 , 273 (2018). [2] S.-C. Yang, Y.-L. Wang, G.-S. Jiao, H.-J. Qian, Z.-Y. Lu, J. Comput. Chem. 37 , 378 (2016). [3] S.-C. Yang, Z.-Y. Lu, H.-J. Qian, Y.-L. Wang, J.-P. Han, Comput. Phys. Commun. 220 , 376 (2017). [ABSTRACT FROM AUTHOR]

To celebrate the leading role Computer Physics Communications (CPC) has played in publishing open-source software in computational physics for over 50 years the editors are delighted to announce this Virtual Special Issue. Since 2018, coinciding with the 50th anniversary of the start of the CPC venture, thirty-two invited articles have been published. Each has been peer reviewed and each bears the header 'CPC 50th anniversary article'. The special issue is in keeping with CPC's ethos: it is focused on computational physics software and is accompanied by twenty-five software systems. The introduction to the collection also includes a personal reflection on Phil Burke, CPC's founder, by Alan Hibbert, a lifelong colleague, who joined Queen's University with Phil in the autumn of 1967. The distinctive feature of CPC is its Program Library which houses and distributes over 3500 open-source programs in computational physics. The introduction concludes with a description of key events in the history of the Program Library, its association with Queen's University Belfast and its transfer to Elsevier's Mendeley Data repository. [ABSTRACT FROM AUTHOR]

diaphane is a portable, scalable, and extensible library for modelling the transport of energy by radiation or relativistic particles (in particular neutrinos). Energy transport modelling is crucial for the hydrodynamic modelling of a wide range of astrophysical phenomena such as planet and galaxy formation, supernova explosions, and cosmic structure evolution. The diaphane library provides a computational framework and functionality to incorporate energy transport modelling into hydrodynamical astrophysics simulations. The transport routines are called from, and operate independently of, the underlying hydrodynamic code of choice. It is designed to be utilized by hydrodynamic astrophysical simulations through a simplified interface layer, allowing it to operate independently of the underlying hydrodynamic code. We provide a sample interface layer to three widely-used astrophysics-focussed smoothed-particle hydrodynamic codes, gadget2 , gasoline and sphynx. The initial release comprises a particle-based implementation of Flux Limited Diffusion and Starrad, a ray casting method that we have developed. diaphane development has been supported the Platform for Advanced Scientific Computing. Program Title: Diaphane Program Files doi: http://dx.doi.org/10.17632/rwyd5k5rjh.1 Code Ocean capsule: https://doi.org/10.24433/CO.5068717.v1 Licensing provisions: MIT Programming language: C/C++ Nature of problem: The ability to model and hence understand many important astrophysical processes has often been limited by the approximations and incompleteness in the treatment of radiation and relativistic particles, even though many algorithms exist for modelling the transport of energy by radiation and relativistic particles. For many problems, multiple algorithms must be used because each algorithm is best-suited in terms of speed and accuracy only in some particular physical regime. A comprehensive library that contains multiple radiation/relativistic transport techniques is needed in order to model astrophysical problems and their wide range in physical regimes. This library framework must be able to be easily and efficiently utilized by a hydrodynamic simulation code. Solution method: We provide a framework for solving radiation and neutrino transport in a portable, scalable, and extensible library. We provide a sample interface for multiple simulation codes, at this time gadget2 [2,3], gasoline [4] and sphynx [5,6]. Community development to develop new algorithms and support new hydrodynamic codes is welcome. The library framework allows multiple energy transport methods to be used simultaneously. Additional comments: This work has been supported by the Swiss Platform for Advanced Scientific Computing (PASC, http://www.pasc-ch.org/) project DIAPHANE (DR and RC) and by the European Research Council (FP7) under ERC Advanced Grant Agreement No. 321263-FISH (RC). A more detailed description of the library and framework are described in [1]. Declaration of Competing Interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. [1] D. S. Reed, T. Dykes, R. Cabezón, C. Gheller, L. Mayer, Computer Physics Communications 226 (2018) 1–9. [2] V. Springel, N. Yoshida, S. D. M. White, New Astronomy6 (2) (2001) 79–117. [3] V. Springel, Mon. Not. R. Astron. Soc. 364 (4) (2005) 1105–1134. [4] J. W. Wadsley, B. W. Keller, T. R. Quinn, Mon. Not. R. Astron. Soc. 471 (2) (2017) 2357–2369. [5] R. M. Cabezon, D. Garcia-Senz, (Sep 2017). [6] R. M. Cabezón, D. García-Senz, J. Figueira, Astronomy & Astrophysics 606 (2017) A78. [ABSTRACT FROM AUTHOR]

This paper presents a version of simulation program, which facilitates the calculation of changes to the intensity of RHEED oscillations in the function of the glancing angle of incidence of the electron beam, employing various models of scattering crystal potential for heteroepitaxial structure of hexagonal (0001) GaN film nucleated on a 3C-SiC surface, including the possible existence of various diffuse scattering models through the layer parallel to the surface. New version program summary Title of program: RHEED_DIFF_Z Program Files doi: http://dx.doi.org/10.17632/52nxmwzkxx.1 Licensing provisions: GNU General Public License 3 Programming language used: C++ Journal reference of previous version: Computer Physics Communications 207 (2016) 536–538 Does the new version supersede the previous version?: No. It is a supplement to the previous version. Reasons for new version: Responding to users’ feedback we present a practical procedure of construction of simulation program, which facilitates the calculation of changes to the intensity of RHEED rocking curves, employing various models of scattering crystal potential for heteroepitaxial structure of hexagonal (0001) GaN film nucleated on a 3C-SiC(111) surface. Nature of problem: Growth of GaN layers on SiC is of importance for potential electronic device applications. Silicon carbide crystallizes in numerous different modifications, so-called polytypes [1]. Among them, only hexagonal structure 4H-SiC and 6H-SiC have been thoroughly investigated while the cubic 3C-SiC polytype is lagging behind in technological developments. The 3C-SiC has a Zincblende crystal structure and is characterized by an identical orientation of all bilayers in the crystal, where the atomic geometry is repeated every three bilayers along the c-axis this crystal structure ( Fig. 1 ). For this reason, 3C-SiC has the high electron mobility and saturation velocity by reducing phonon scattering resulting from the high symmetry. The lattice mismatch between GaN(0001) and 3C-SiC(111) is about 3.3%, and is smaller than the mismatch between GaN(0001) and Si(111) [2]. Therefore progress in investigations on crystal growth of GaN on 3C-SiC is a key issue for device developments related to this heterostructure. Researchers and technologists manufacturing two-dimensional heterostructures frequently use RHEED rocking curves to control growth of samples at the atomistic level of accuracy. The fundamental scientific problem in such research is to specify both interface type and growth mechanism for subsequent layers. Method of solution: RHEED intensities are calculated within the general framework described in Refs. [2] and [3]. Summary of revisions: The presented version of the program implements an original algorithm of self-consistent calculations for scattering potentials GaN(0001)/3C-SiC(111) and solving a time-independent Schrödinger equation for high-energy electrons. During the numerical calculations of the changes in the intensity of the specular beam in the function of the glancing angle, it was assumed that the azimuth of the incident beam direction corresponds to the one-beam condition, the electron energy equals 10 keV, the glancing angle was increased from 0 . 5 ∘ to 6 . 5 ∘ , and the value of α parameter = 0 . 1 and β = 0 . 5 for the MODEL3 of the scattering potential [3] with the model of ideal SiC/GaN interface [4]. References [1] G. R. Fisher and P. Barnes, Phil. Mag. B 61(2) (1990) 217-236. [2] A. Daniluk, Comput. Phys. Commun. 207 (2016) 536-538. [3] A. Daniluk, Comput. Phys. Commun. 185 (2014) 3001-3009. [4] H. Morkoç, Handbook of Nitride Semiconductors and Devices, Vol. 1: Materials Properties, Physics and Growth, Wiley VCH 2008. [ABSTRACT FROM AUTHOR]