Computer Physics Communications最新文献_第3页

OpenSBLI v3.0: High-fidelity multi-block transonic aerofoil CFD simulations using domain specific languages on GPUs OpenSBLI v3.0：在 GPU 上使用特定领域语言进行高保真多区块跨声速气流箔 CFD 仿真

IF 7.2 2区物理与天体物理 Q1 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Computer Physics Communications

Pub Date : 2024-10-24 DOI: 10.1016/j.cpc.2024.109406

David J. Lusher , Andrea Sansica , Neil D. Sandham , Jianping Meng , Bálint Siklósi , Atsushi Hashimoto

OpenSBLI is an automatic code-generation framework for compressible Computational Fluid Dynamics (CFD) simulations on heterogeneous computing architectures (previous release: Lusher et al. (2021) [4]). OpenSBLI is coupled to the Oxford Parallel Structured (OPS) Domain Specific Language (DSL), which uses source-to-source translation to enable parallel execution of the code on large-scale supercomputers, including multi-GPU clusters. To date, OpenSBLI has largely been applied to compressible turbulence and shock-wave/boundary-layer interactions on very simple geometries comprised of single mesh blocks with essentially orthogonal grid lines. OpenSBLI has been extended in this new release to target strongly curvilinear cases, including transonic aerofoils using multi-block grids. In addition to multi-block mesh support, more efficient numerical shock-capturing methods and filters have been added to the codebase. Improvements to post-processing, reduced-dimension data output, and coupling to a modal decomposition library are also included. A set of validation cases are presented to showcase the new code features. Furthermore, state-of-the-art wide-span transonic aerofoil simulations on up to

N = 2.5 \times 10^{9}

grid points demonstrate that wider aspect ratios can alter buffet predictions and increase the regularity of the low-frequency shock oscillations by accommodating fully-developed trailing edge flow separation. Spectral Proper Orthogonal Decomposition (SPOD) analysis showed that overly-narrow aerofoil simulations contain additional domain-dependent energy content at a Strouhal number of

S t \approx 3

associated with wake modes.

OpenSBLI 是一个自动代码生成框架，用于在异构计算架构上进行可压缩计算流体动力学（CFD）模拟（上一版本：Lusher 等人 (2021) [4]）：Lusher et al. (2021) [4]）。OpenSBLI与牛津并行结构化（OPS）领域专用语言（DSL）相耦合，后者使用源对源翻译，使代码能够在大规模超级计算机（包括多GPU集群）上并行执行。迄今为止，OpenSBLI 主要应用于可压缩湍流和冲击波/边界层相互作用，其几何结构非常简单，由基本正交网格线的单一网格块组成。OpenSBLI 在新版本中进行了扩展，可用于强曲线情况，包括使用多块网格的跨音速气膜。除了支持多块网格外，代码库中还添加了更高效的数值冲击捕获方法和滤波器。此外，还改进了后处理、缩减维度数据输出以及与模态分解库的耦合。为展示新的代码功能，还提供了一组验证案例。此外，在多达 N=2.5×109 个网格点上进行的最先进的大跨度跨音速气膜模拟表明，更宽的纵横比可以改变缓冲预测，并通过容纳充分发展的后缘流分离来增加低频冲击振荡的规律性。光谱适当正交分解（SPOD）分析表明，过窄的气膜模拟在斯特劳哈尔数为 St≈3 时包含与尾流模式相关的额外域相关能量含量。

{"title":"OpenSBLI v3.0: High-fidelity multi-block transonic aerofoil CFD simulations using domain specific languages on GPUs","authors":"David J. Lusher , Andrea Sansica , Neil D. Sandham , Jianping Meng , Bálint Siklósi , Atsushi Hashimoto","doi":"10.1016/j.cpc.2024.109406","DOIUrl":"10.1016/j.cpc.2024.109406","url":null,"abstract":"<div><div>OpenSBLI is an automatic code-generation framework for compressible Computational Fluid Dynamics (CFD) simulations on heterogeneous computing architectures (previous release: Lusher et al. (2021) <span><span>[4]</span></span>). OpenSBLI is coupled to the Oxford Parallel Structured (OPS) Domain Specific Language (DSL), which uses source-to-source translation to enable parallel execution of the code on large-scale supercomputers, including multi-GPU clusters. To date, OpenSBLI has largely been applied to compressible turbulence and shock-wave/boundary-layer interactions on very simple geometries comprised of single mesh blocks with essentially orthogonal grid lines. OpenSBLI has been extended in this new release to target strongly curvilinear cases, including transonic aerofoils using multi-block grids. In addition to multi-block mesh support, more efficient numerical shock-capturing methods and filters have been added to the codebase. Improvements to post-processing, reduced-dimension data output, and coupling to a modal decomposition library are also included. A set of validation cases are presented to showcase the new code features. Furthermore, state-of-the-art wide-span transonic aerofoil simulations on up to <span><math><mi>N</mi><mo>=</mo><mn>2.5</mn><mo>×</mo><msup><mrow><mn>10</mn></mrow><mrow><mn>9</mn></mrow></msup></math></span> grid points demonstrate that wider aspect ratios can alter buffet predictions and increase the regularity of the low-frequency shock oscillations by accommodating fully-developed trailing edge flow separation. Spectral Proper Orthogonal Decomposition (SPOD) analysis showed that overly-narrow aerofoil simulations contain additional domain-dependent energy content at a Strouhal number of <span><math><mi>S</mi><mi>t</mi><mo>≈</mo><mn>3</mn></math></span> associated with wake modes.</div></div>","PeriodicalId":285,"journal":{"name":"Computer Physics Communications","volume":"307 ","pages":"Article 109406"},"PeriodicalIF":7.2,"publicationDate":"2024-10-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142554022","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"物理与天体物理","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

An adaptive learning strategy for surrogate modeling of high-dimensional functions - Application to unsteady hypersonic flows in chemical nonequilibrium 高维函数代理建模的自适应学习策略--应用于化学非平衡态下的非稳态高超声速流动

IF 7.2 2区物理与天体物理 Q1 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Computer Physics Communications

Pub Date : 2024-10-24 DOI: 10.1016/j.cpc.2024.109404

Clément Scherding , Georgios Rigas , Denis Sipp , Peter J. Schmid , Taraneh Sayadi

Many engineering applications rely on the evaluation of expensive, non-linear high-dimensional functions. In this paper, we propose the RONAALP algorithm (Reduced Order Nonlinear Approximation with Adaptive Learning Procedure) to incrementally learn, on-the-fly as the application progresses, a fast and accurate reduced-order surrogate model of a target function. First, a combination of nonlinear auto-encoder, community clustering, and radial basis function networks allows us to learn an efficient and compact surrogate model with limited training data. Secondly, an active learning procedure overcomes any extrapolation issues during the online stage by adapting the surrogate model with high-fidelity evaluations that fall outside its current validity range. This approach results in generalizable, fast, and accurate reduced-order models of high-dimensional functions. The method is demonstrated on three direct numerical simulations of hypersonic flows in chemical nonequilibrium. Accurate simulations of these flows rely on detailed thermochemical gas models that dramatically increase the cost of such calculations. Using RONAALP to learn a reduced-order thermodynamic model surrogate on-the-fly, the cost of such simulations was reduced by up to 75% while maintaining an error of less than 10% on relevant quantities of interest.

许多工程应用依赖于对昂贵的非线性高维函数进行评估。在本文中，我们提出了 RONAALP 算法（Reduced Order Nonlinear Approximation with Adaptive Learning Procedure，自适应学习程序的降阶非线性逼近算法），随着应用的进展，逐步学习目标函数的快速、准确的降阶代理模型。首先，结合非线性自动编码器、群集聚类和径向基函数网络，我们可以利用有限的训练数据学习高效、紧凑的代用模型。其次，主动学习程序克服了在线阶段的任何外推问题，通过对当前有效范围之外的高保真评估来调整代理模型。通过这种方法，可以建立可通用、快速、准确的高维函数降阶模型。该方法在化学非平衡态高超音速流动的三次直接数值模拟中得到了验证。对这些流动的精确模拟依赖于详细的热化学气体模型，这大大增加了计算成本。使用 RONAALP 实时学习减阶热力学代用模型，此类模拟的成本最多可降低 75%，而相关相关量的误差却保持在 10%以下。

{"title":"An adaptive learning strategy for surrogate modeling of high-dimensional functions - Application to unsteady hypersonic flows in chemical nonequilibrium","authors":"Clément Scherding , Georgios Rigas , Denis Sipp , Peter J. Schmid , Taraneh Sayadi","doi":"10.1016/j.cpc.2024.109404","DOIUrl":"10.1016/j.cpc.2024.109404","url":null,"abstract":"<div><div>Many engineering applications rely on the evaluation of expensive, non-linear high-dimensional functions. In this paper, we propose the RONAALP algorithm (<strong>R</strong>educed <strong>O</strong>rder <strong>N</strong>onlinear <strong>A</strong>pproximation with <strong>A</strong>daptive <strong>L</strong>earning <strong>P</strong>rocedure) to incrementally learn, on-the-fly as the application progresses, a fast and accurate reduced-order surrogate model of a target function. First, a combination of nonlinear auto-encoder, community clustering, and radial basis function networks allows us to learn an efficient and compact surrogate model with limited training data. Secondly, an active learning procedure overcomes any extrapolation issues during the online stage by adapting the surrogate model with high-fidelity evaluations that fall outside its current validity range. This approach results in generalizable, fast, and accurate reduced-order models of high-dimensional functions. The method is demonstrated on three direct numerical simulations of hypersonic flows in chemical nonequilibrium. Accurate simulations of these flows rely on detailed thermochemical gas models that dramatically increase the cost of such calculations. Using RONAALP to learn a reduced-order thermodynamic model surrogate on-the-fly, the cost of such simulations was reduced by up to 75% while maintaining an error of less than 10% on relevant quantities of interest.</div></div>","PeriodicalId":285,"journal":{"name":"Computer Physics Communications","volume":"307 ","pages":"Article 109404"},"PeriodicalIF":7.2,"publicationDate":"2024-10-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142561012","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"物理与天体物理","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

MCEND: An open-source program for quantum electron-nuclear dynamics MCEND：量子电子核动力学开源程序

IF 7.2 2区物理与天体物理 Q1 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Computer Physics Communications

Pub Date : 2024-10-23 DOI: 10.1016/j.cpc.2024.109405

Inga S. Ulusoy , Lucas E. Aebersold , Cong Wang , Angela K. Wilson

The software MCEND (Multi-Configuration Electron-Nuclear Dynamics) is a free open-source program package which simulates the quantum dynamics of electron-nuclei simultaneously for diatomic molecules. Its formulation, implementation, and usage are described in detail. MCEND uses a grid-based basis representation for the nuclei, and the electronic basis is derived from standard electronic structure basis sets on the nuclear grid. The wave function is represented as a sum over products of electronic and nuclear wave functions, thus capturing correlation effects between electrons, nuclei, and electrons and nuclei. The LiH molecule was used as an example for simulating the molecular properties such as the dipole moment and absorption spectrum.

PROGRAM SUMMARY

Program Title MCEND, v.2.6.0

CPC Library link to program files: https://doi.org/10.17632/tkb9dwf85t.1

Developer's repository link: https://github.com/MCEND-hub/MCEND (https://github.com/MCEND-hub/MCEND-library and https://github.com/MCEND-hub/MCEND-tools are git submodules of MCEND)

Licensing provisions: MIT

Programming language: Fortran 90 and Python 3

External routines/libraries: FFTW, OpenMP, BLAS, LAPACK, PSI4, Matplotlib, mendeleev, NumPy, Pandas, SciPy, PyTables

Nature of problem: MCEND is to simulate the quantum dynamics of electrons and nuclei simultaneously at multiconfiguration levels.

Solution method: The presented program package solves the time-dependent Schrödinger equation with the wave function represented as sum over configuration products using an 8th-order adaptive step size Runge-Kutta ordinary differential equation (ODE) solver. The software can be extended by supplementing modules on the existing infrastructure.

MCEND（Multi-Configuration Electron-Nuclear Dynamics）软件是一个免费的开放源码程序包，可同时模拟二原子分子的电子核量子动力学。本文详细介绍了该软件的开发、实施和使用。MCEND 采用基于网格的原子核基础表示法，电子基础来自核网格上的标准电子结构基础集。波函数表示为电子波函数与核波函数乘积之和，从而捕捉电子、原子核以及电子与原子核之间的相关效应。以 LiH 分子为例，模拟了偶极矩和吸收光谱等分子特性。PROGRAM SUMMARYProgram Title MCEND, v.2.6.0CPC Library link to program files: https://doi.org/10.17632/tkb9dwf85t.1Developer's repository link: https://github.com/MCEND-hub/MCEND (https://github.com/MCEND-hub/MCEND-library and https://github.com/MCEND-hub/MCEND-tools are git submodules of MCEND)Licensing provisions：MIT 编程语言：Fortran 90 和 Python 3Fortran 90 和 Python 3外部例程/库：FFTW、OpenMP、BLAS、LAPACK、PSI4、Matplotlib、mendeleev、NumPy、Pandas、SciPy、PyTables问题性质：MCEND 是在多配置水平上同时模拟电子和原子核的量子动力学：所介绍的程序包使用 8 阶自适应步长 Runge-Kutta 常微分方程求解器求解与时间相关的薛定谔方程，波函数表示为配置乘积之和。该软件可通过在现有基础架构上补充模块进行扩展。

{"title":"MCEND: An open-source program for quantum electron-nuclear dynamics","authors":"Inga S. Ulusoy , Lucas E. Aebersold , Cong Wang , Angela K. Wilson","doi":"10.1016/j.cpc.2024.109405","DOIUrl":"10.1016/j.cpc.2024.109405","url":null,"abstract":"<div><div>The software MCEND (Multi-Configuration Electron-Nuclear Dynamics) is a free open-source program package which simulates the quantum dynamics of electron-nuclei simultaneously for diatomic molecules. Its formulation, implementation, and usage are described in detail. MCEND uses a grid-based basis representation for the nuclei, and the electronic basis is derived from standard electronic structure basis sets on the nuclear grid. The wave function is represented as a sum over products of electronic and nuclear wave functions, thus capturing correlation effects between electrons, nuclei, and electrons and nuclei. The LiH molecule was used as an example for simulating the molecular properties such as the dipole moment and absorption spectrum.</div><div><strong>PROGRAM SUMMARY</strong></div><div><em>Program Title</em> MCEND, v.2.6.0</div><div><em>CPC Library link to program files:</em> <span><span>https://doi.org/10.17632/tkb9dwf85t.1</span><svg><path></path></svg></span></div><div><em>Developer's repository link:</em> <span><span>https://github.com/MCEND-hub/MCEND</span><svg><path></path></svg></span> (<span><span>https://github.com/MCEND-hub/MCEND-library</span><svg><path></path></svg></span> and <span><span>https://github.com/MCEND-hub/MCEND-tools are git submodules of MCEND</span><svg><path></path></svg></span>)</div><div><em>Licensing provisions:</em> MIT</div><div><em>Programming language:</em> Fortran 90 and Python 3</div><div><em>External routines/libraries:</em> FFTW, OpenMP, BLAS, LAPACK, PSI4, Matplotlib, mendeleev, NumPy, Pandas, SciPy, PyTables</div><div><em>Nature of problem:</em> MCEND is to simulate the quantum dynamics of electrons and nuclei simultaneously at multiconfiguration levels.</div><div><em>Solution method:</em> The presented program package solves the time-dependent Schrödinger equation with the wave function represented as sum over configuration products using an 8th-order adaptive step size Runge-Kutta ordinary differential equation (ODE) solver. The software can be extended by supplementing modules on the existing infrastructure.</div></div>","PeriodicalId":285,"journal":{"name":"Computer Physics Communications","volume":"307 ","pages":"Article 109405"},"PeriodicalIF":7.2,"publicationDate":"2024-10-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142554027","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"物理与天体物理","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

VAN-DAMME: GPU-accelerated and symmetry-assisted quantum optimal control of multi-qubit systems VAN-DAMME：多量子比特系统的 GPU 加速和对称辅助量子优化控制

IF 7.2 2区物理与天体物理 Q1 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Computer Physics Communications

Pub Date : 2024-10-17 DOI: 10.1016/j.cpc.2024.109403

José M. Rodríguez-Borbón , Xian Wang , Adrián P. Diéguez , Khaled Z. Ibrahim , Bryan M. Wong

We present an open-source software package, VAN-DAMME (Versatile Approaches to Numerically Design, Accelerate, and Manipulate Magnetic Excitations), for massively-parallelized quantum optimal control (QOC) calculations of multi-qubit systems. To enable large QOC calculations, the VAN-DAMME software package utilizes symmetry-based techniques with custom GPU-enhanced algorithms. This combined approach allows for the simultaneous computation of hundreds of matrix exponential propagators that efficiently leverage the intra-GPU parallelism found in high-performance GPUs. In addition, to maximize the computational efficiency of the VAN-DAMME code, we carried out several extensive tests on data layout, computational complexity, memory requirements, and performance. These extensive analyses allowed us to develop computationally efficient approaches for evaluating complex-valued matrix exponential propagators based on Padé approximants. To assess the computational performance of our GPU-accelerated VAN-DAMME code, we carried out QOC calculations of systems containing 10 - 15 qubits, which showed that our GPU implementation is 18.4× faster than the corresponding CPU implementation. Our GPU-accelerated enhancements allow efficient calculations of multi-qubit systems, which can be used for the efficient implementation of QOC applications across multiple domains.

Program summary

Program Title: VAN-DAMME

CPC Library link to program files:: https://doi.org/10.17632/zcgw2n5bjf.1

Licensing provisions: GNU General Public License 3

Programming language: C++ and CUDA

Nature of problem: The VAN-DAMME software package utilizes GPU-accelerated routines and new algorithmic improvements to compute optimized time-dependent magnetic fields that can drive a system from a known initial qubit configuration to a specified target state with a large (≈1) transition probability.

Solution method: Quantum control, GPU acceleration, analytic gradients, matrix exponential, and gradient ascent optimization.

我们介绍了一个开源软件包 VAN-DAMME（数值设计、加速和操纵磁激发的多功能方法），用于多量子比特系统的大规模并行量子优化控制（QOC）计算。为了实现大型 QOC 计算，VAN-DAMME 软件包利用了基于对称性的技术和定制的 GPU 增强算法。这种组合方法可以同时计算数百个矩阵指数传播者，有效利用高性能 GPU 中的 GPU 内部并行性。此外，为了最大限度地提高 VAN-DAMME 代码的计算效率，我们对数据布局、计算复杂性、内存需求和性能进行了多次广泛测试。通过这些广泛的分析，我们开发出了计算效率高的方法，用于评估基于帕代近似值的复值矩阵指数传播者。为了评估我们的GPU加速VAN-DAMME代码的计算性能，我们对包含10-15个量子比特的系统进行了QOC计算，结果表明我们的GPU实现比相应的CPU实现快18.4倍。我们的GPU加速增强技术可以实现多量子比特系统的高效计算，可用于跨多个领域的QOC应用的高效实施：VAN-DAMMECPC 程序库链接到程序文件：: https://doi.org/10.17632/zcgw2n5bjf.1Licensing provisions：GNU General Public License 3编程语言：问题性质：VAN-DAMME 软件包利用 GPU 加速例程和新的算法改进来计算优化的随时间变化的磁场，该磁场可以驱动一个系统从已知的初始量子比特配置以大（≈1）的过渡概率到达指定的目标状态：量子控制、GPU 加速、解析梯度、矩阵指数和梯度上升优化。

{"title":"VAN-DAMME: GPU-accelerated and symmetry-assisted quantum optimal control of multi-qubit systems","authors":"José M. Rodríguez-Borbón , Xian Wang , Adrián P. Diéguez , Khaled Z. Ibrahim , Bryan M. Wong","doi":"10.1016/j.cpc.2024.109403","DOIUrl":"10.1016/j.cpc.2024.109403","url":null,"abstract":"<div><div>We present an open-source software package, VAN-DAMME (Versatile Approaches to Numerically Design, Accelerate, and Manipulate Magnetic Excitations), for massively-parallelized quantum optimal control (QOC) calculations of multi-qubit systems. To enable large QOC calculations, the VAN-DAMME software package utilizes symmetry-based techniques with custom GPU-enhanced algorithms. This combined approach allows for the simultaneous computation of hundreds of matrix exponential propagators that efficiently leverage the intra-GPU parallelism found in high-performance GPUs. In addition, to maximize the computational efficiency of the VAN-DAMME code, we carried out several extensive tests on data layout, computational complexity, memory requirements, and performance. These extensive analyses allowed us to develop computationally efficient approaches for evaluating complex-valued matrix exponential propagators based on Padé approximants. To assess the computational performance of our GPU-accelerated VAN-DAMME code, we carried out QOC calculations of systems containing 10 - 15 qubits, which showed that our GPU implementation is 18.4× faster than the corresponding CPU implementation. Our GPU-accelerated enhancements allow efficient calculations of multi-qubit systems, which can be used for the efficient implementation of QOC applications across multiple domains.</div></div><div><h3>Program summary</h3><div><em>Program Title:</em> VAN-DAMME</div><div><em>CPC Library link to program files::</em> <span><span>https://doi.org/10.17632/zcgw2n5bjf.1</span><svg><path></path></svg></span></div><div><em>Licensing provisions:</em> GNU General Public License 3</div><div><em>Programming language:</em> C++ and CUDA</div><div><em>Nature of problem:</em> The VAN-DAMME software package utilizes GPU-accelerated routines and new algorithmic improvements to compute optimized time-dependent magnetic fields that can drive a system from a known initial qubit configuration to a specified target state with a large (≈1) transition probability.</div><div><em>Solution method:</em> Quantum control, GPU acceleration, analytic gradients, matrix exponential, and gradient ascent optimization.</div></div>","PeriodicalId":285,"journal":{"name":"Computer Physics Communications","volume":"307 ","pages":"Article 109403"},"PeriodicalIF":7.2,"publicationDate":"2024-10-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142534983","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"物理与天体物理","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

The FLUKA Monte Carlo simulation of the magnetic spectrometer of the FOOT experiment FOOT 实验磁谱仪的 FLUKA 蒙特卡洛模拟

IF 7.2 2区物理与天体物理 Q1 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Computer Physics Communications

Pub Date : 2024-10-16 DOI: 10.1016/j.cpc.2024.109398

Y. Dong , S.M. Valle , G. Battistoni , I. Mattei , C. Finck , V. Patera , A. Alexandrov , B. Alpat , G. Ambrosi , S. Argirò , M. Barbanera , N. Bartosik , M.G. Bisogni , V. Boccia , F. Cavanna , P. Cerello , E. Ciarrocchi , A. De Gregorio , G. De Lellis , A. Di Crescenzo , S. Muraro

The FOOT experiment of INFN is devoted to the measurement of the nuclear fragmentation double differential cross sections useful for the improvement of calculation models adopted in hadrontherapy and radioprotection. A detailed Monte Carlo simulation of the FOOT magnetic spectrometer has been implemented in order to optimize the design and to guide data analysis. This task has been accomplished by means of the FLUKA Monte Carlo code. The input files of the FLUKA simulations are created from the software framework of the experiment, in order to have a consistent generation and description of geometry and materials in both simulation and data analysis. In addition, this ensures the possibility of processing both simulated and real data with the same data analysis procedures. Databases containing specific parameters describing the setup employed in each different data taking campaign are used. A customized event-by-event output of the Monte Carlo code has been developed. It can be read out by the general software framework of FOOT, enabling access to the generation history of all particles in the same event. This output structure therefore gives the possibility to perform a detailed analysis and study of all relevant processes, allowing the detailed tracking reconstruction of all individual particles. Examples of results are presented.

INFN 的 FOOT 实验致力于测量核碎裂双差分截面，这对改进放射性治疗和放射防护中采用的计算模型非常有用。为了优化设计和指导数据分析，对 FOOT 磁谱仪进行了详细的蒙特卡罗模拟。这项任务是通过 FLUKA 蒙特卡罗代码完成的。FLUKA 模拟的输入文件是根据实验的软件框架创建的，以便在模拟和数据分析中对几何形状和材料进行一致的生成和描述。此外，这还确保了使用相同的数据分析程序处理模拟数据和真实数据的可能性。数据库中包含描述每个不同数据采集活动所使用的设置的特定参数。蒙地卡罗代码的逐个事件定制输出已经开发出来。它可以由 FOOT 的通用软件框架读出，从而能够访问同一事件中所有粒子的生成历史。因此，这种输出结构提供了对所有相关过程进行详细分析和研究的可能性，允许对所有单个粒子进行详细的跟踪重建。现介绍一些结果实例。

{"title":"The FLUKA Monte Carlo simulation of the magnetic spectrometer of the FOOT experiment","authors":"Y. Dong , S.M. Valle , G. Battistoni , I. Mattei , C. Finck , V. Patera , A. Alexandrov , B. Alpat , G. Ambrosi , S. Argirò , M. Barbanera , N. Bartosik , M.G. Bisogni , V. Boccia , F. Cavanna , P. Cerello , E. Ciarrocchi , A. De Gregorio , G. De Lellis , A. Di Crescenzo , S. Muraro","doi":"10.1016/j.cpc.2024.109398","DOIUrl":"10.1016/j.cpc.2024.109398","url":null,"abstract":"<div><div>The FOOT experiment of INFN is devoted to the measurement of the nuclear fragmentation double differential cross sections useful for the improvement of calculation models adopted in hadrontherapy and radioprotection. A detailed Monte Carlo simulation of the FOOT magnetic spectrometer has been implemented in order to optimize the design and to guide data analysis. This task has been accomplished by means of the FLUKA Monte Carlo code. The input files of the FLUKA simulations are created from the software framework of the experiment, in order to have a consistent generation and description of geometry and materials in both simulation and data analysis. In addition, this ensures the possibility of processing both simulated and real data with the same data analysis procedures. Databases containing specific parameters describing the setup employed in each different data taking campaign are used. A customized event-by-event output of the Monte Carlo code has been developed. It can be read out by the general software framework of FOOT, enabling access to the generation history of all particles in the same event. This output structure therefore gives the possibility to perform a detailed analysis and study of all relevant processes, allowing the detailed tracking reconstruction of all individual particles. Examples of results are presented.</div></div>","PeriodicalId":285,"journal":{"name":"Computer Physics Communications","volume":"307 ","pages":"Article 109398"},"PeriodicalIF":7.2,"publicationDate":"2024-10-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142535613","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"物理与天体物理","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Symmetry adaptation for self-consistent many-body calculations 自洽多体计算的对称性调整

IF 7.2 2区物理与天体物理 Q1 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Computer Physics Communications

Pub Date : 2024-10-16 DOI: 10.1016/j.cpc.2024.109401

Xinyang Dong , Emanuel Gull

The exploitation of space group symmetries in numerical calculations of periodic crystalline solids accelerates calculations and provides physical insight. We present results for a space-group symmetry adaptation of electronic structure calculations within the finite-temperature self-consistent GW method along with an efficient parallelization scheme on accelerators. Our implementation employs the simultaneous diagonalization of the Dirac characters of the orbital representation. Results show that symmetry adaptation in self-consistent many-body codes results in substantial improvements of the runtime, and that block diagonalization on top of a restriction to the irreducible wedge results in additional speedup.

在周期性晶体固体的数值计算中利用空间群对称性可加快计算速度并提供物理洞察力。我们介绍了在有限温度自洽 GW 方法中对电子结构计算进行空间群对称性调整的结果，以及在加速器上的高效并行化方案。我们的实现方法采用了轨道表示的狄拉克特征的同时对角化。结果表明，自洽多体代码中的对称性适应可大幅改善运行时间，而在限制不可还原楔的基础上进行分块对角化可进一步提速。

引用次数: 0

Stochastic automatic differentiation for Monte Carlo processes 蒙特卡罗过程的随机自动微分

IF 7.2 2区物理与天体物理 Q1 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Computer Physics Communications

Pub Date : 2024-10-15 DOI: 10.1016/j.cpc.2024.109396

Guilherme Catumba, Alberto Ramos, Bryan Zaldivar

Monte Carlo methods represent a cornerstone of computer science. They allow sampling high dimensional distribution functions in an efficient way. In this paper we consider the extension of Automatic Differentiation (AD) techniques to Monte Carlo processes, addressing the problem of obtaining derivatives (and in general, the Taylor series) of expectation values. Borrowing ideas from the lattice field theory community, we examine two approaches. One is based on reweighting while the other represents an extension of the Hamiltonian approach typically used by the Hybrid Monte Carlo (HMC) and similar algorithms. We show that the Hamiltonian approach can be understood as a change of variables of the reweighting approach, resulting in much reduced variances of the coefficients of the Taylor series. This work opens the door to finding other variance reduction techniques for derivatives of expectation values.

蒙特卡罗方法是计算机科学的基石。它们允许以高效的方式对高维分布函数进行采样。在本文中，我们考虑将自动微分（Automatic Differentiation，AD）技术扩展到蒙特卡洛过程，解决获取期望值导数（以及一般情况下的泰勒级数）的问题。我们借鉴格子场理论界的观点，研究了两种方法。一种是基于重新加权，另一种是对混合蒙特卡罗（HMC）和类似算法通常使用的汉密尔顿方法的扩展。我们表明，哈密顿方法可以理解为重新加权方法的变量变化，从而大大降低了泰勒级数系数的方差。这项工作为寻找其他降低期望值导数方差的技术打开了大门。

引用次数: 0

LinApart: Optimizing the univariate partial fraction decomposition LinApart：优化单变量部分分数分解

IF 7.2 2区物理与天体物理 Q1 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Computer Physics Communications

Pub Date : 2024-10-15 DOI: 10.1016/j.cpc.2024.109395

B. Chargeishvili , L. Fekésházy , G. Somogyi , S. Van Thurenhout

We present LinApart, a routine designed for efficiently performing the univariate partial fraction decomposition of large symbolic expressions. Our method is based on an explicit closed formula for the decomposition of rational functions with fully factorized denominators. We provide an implementation in the Wolfram Mathematica language, which can lead to very significant performance gains over the built-in Apart command. Furthermore, a C language library implementing the core functionality and suitable for interfacing with other software is also provided. Both codes are made available at https://github.com/fekeshazy/LinApart.

我们介绍的 LinApart 是一种用于高效地对大型符号表达式进行单变量部分分数分解的例程。我们的方法基于有理函数分解的显式封闭公式，其分母完全因式分解。我们提供了 Wolfram Mathematica 语言的实现方法，与内置的 Apart 命令相比，该方法的性能提升非常明显。此外，我们还提供了一个实现核心功能的 C 语言库，适合与其他软件连接。这两个代码都可在 https://github.com/fekeshazy/LinApart 上获得。

引用次数: 0

A unified machine learning approach for reconstructing hadronically decaying tau leptons 重构强子衰变头轻子的统一机器学习方法

IF 7.2 2区物理与天体物理 Q1 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Computer Physics Communications

Pub Date : 2024-10-15 DOI: 10.1016/j.cpc.2024.109399

Laurits Tani , Nalong-Norman Seeba , Hardi Vanaveski , Joosep Pata , Torben Lange

Tau leptons serve as an important tool for studying the production of Higgs and electroweak bosons, both within and beyond the Standard Model of particle physics. Accurate reconstruction and identification of hadronically decaying tau leptons is a crucial task for current and future high energy physics experiments. Given the advances in jet tagging, we demonstrate how tau lepton reconstruction can be decomposed into tau identification, kinematic reconstruction, and decay mode classification in a multi-task machine learning setup. Based on an electron-positron collision dataset with full detector simulation and reconstruction, we show that common jet tagging architectures can be effectively used for these sub-tasks. We achieve comparable momentum resolutions of 2–3% with all the tested models, while the precision of reconstructing individual decay modes is between 80–95%. We find ParticleTransformer to be the best-performing approach, significantly outperforming the heuristic baseline. This paper also serves as an introduction to a new publicly available Fuτure dataset for the development of tau reconstruction algorithms. This allows to further study the resilience of ML models to domain shifts and the efficient use of foundation models for such tasks.

头轻子是研究希格斯玻色子和电弱玻色子产生的重要工具，无论是在粒子物理标准模型之内还是之外。精确重建和识别强子衰变的头轻子是当前和未来高能物理实验的关键任务。鉴于射流标签技术的进步，我们展示了如何在多任务机器学习设置中将头轻子重构分解为头识别、运动学重构和衰变模式分类。基于电子-正电子碰撞数据集的完整探测器模拟和重构，我们展示了常见的射流标记架构可以有效地用于这些子任务。所有测试模型的动量分辨率都在 2-3% 之间，而重建单个衰变模式的精度则在 80-95% 之间。我们发现粒子转换器（ParticleTransformer）是表现最好的方法，明显优于启发式基线。本文还介绍了用于开发 tau 重建算法的新的公开 Fuτure 数据集。这有助于进一步研究 ML 模型对领域变化的适应能力，以及在此类任务中对基础模型的有效使用。

{"title":"A unified machine learning approach for reconstructing hadronically decaying tau leptons","authors":"Laurits Tani , Nalong-Norman Seeba , Hardi Vanaveski , Joosep Pata , Torben Lange","doi":"10.1016/j.cpc.2024.109399","DOIUrl":"10.1016/j.cpc.2024.109399","url":null,"abstract":"<div><div>Tau leptons serve as an important tool for studying the production of Higgs and electroweak bosons, both within and beyond the Standard Model of particle physics. Accurate reconstruction and identification of hadronically decaying tau leptons is a crucial task for current and future high energy physics experiments. Given the advances in jet tagging, we demonstrate how tau lepton reconstruction can be decomposed into tau identification, kinematic reconstruction, and decay mode classification in a multi-task machine learning setup. Based on an electron-positron collision dataset with full detector simulation and reconstruction, we show that common jet tagging architectures can be effectively used for these sub-tasks. We achieve comparable momentum resolutions of 2–3% with all the tested models, while the precision of reconstructing individual decay modes is between 80–95%. We find ParticleTransformer to be the best-performing approach, significantly outperforming the heuristic baseline. This paper also serves as an introduction to a new publicly available <span>Fu</span><em>τ</em><span>ure</span> dataset for the development of tau reconstruction algorithms. This allows to further study the resilience of ML models to domain shifts and the efficient use of foundation models for such tasks.</div></div>","PeriodicalId":285,"journal":{"name":"Computer Physics Communications","volume":"307 ","pages":"Article 109399"},"PeriodicalIF":7.2,"publicationDate":"2024-10-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142438238","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"物理与天体物理","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Neural-Parareal: Self-improving acceleration of fusion MHD simulations using time-parallelisation and neural operators 神经-并行：利用时间并行化和神经算子对融合 MHD 仿真进行自我改进加速

IF 7.2 2区物理与天体物理 Q1 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Computer Physics Communications

Pub Date : 2024-10-15 DOI: 10.1016/j.cpc.2024.109391

S.J.P. Pamela , N. Carey , J. Brandstetter , R. Akers , L. Zanisi , J. Buchanan , V. Gopakumar , M. Hoelzl , G. Huijsmans , K. Pentland , T. James , G. Antonucci , JOREK Team

The fusion research facility ITER is currently being assembled to demonstrate that fusion can be used for industrial energy production, while several other programmes across the world are also moving forward, such as EU-DEMO, CFETR, SPARC and STEP. The high engineering complexity of a tokamak makes it an extremely challenging device to optimise, and test-based optimisation would be too slow and too costly. Instead, digital design and optimisation must be favoured, which requires strongly-coupled suites of multi-physics, multi-scale High-Performance Computing calculations. Safety regulation, uncertainty quantification, and optimisation of fusion digital twins is undoubtedly an Exascale grand challenge. In this context, having surrogate models to provide quick estimates with uncertainty quantification is essential to explore and optimise new design options. But there lies the dilemma: accurate surrogate training first requires simulation data. Extensive work has explored solver-in-the-loop solutions to maximise the training of such surrogates. Likewise, innovative methods have been proposed to accelerate conventional HPC solvers using surrogates. Here, a novel approach is designed to do both. By bootstrapping neural operators and HPC methods together, a self-improving framework is achieved. As more simulations are being run within the framework, the surrogate improves, while the HPC simulations get accelerated. This idea is demonstrated on fusion-relevant MHD simulations, where Fourier Neural Operator based surrogates are used to create neural coarse-solver for the Parareal (time-parallelisation) method. Parareal is particularly relevant for large HPC simulations where conventional spatial parallelisation has saturated, and the temporal dimension is thus parallelised as well. This Neural-Parareal framework is a step towards exploiting the convergence of HPC and AI, where scientists and engineers can benefit from automated, self-improving, ever faster simulations. All data/codes developed here are made available to the community.

目前正在组装国际热核聚变实验堆（ITER）聚变研究设施，以证明聚变可用于工业能源生产，同时全球其他几个计划也在推进之中，如欧盟-DEMO、CFETR、SPARC 和 STEP。托卡马克的高工程复杂性使其成为一个极具挑战性的优化设备，而基于测试的优化方法速度太慢、成本太高。因此，必须采用数字设计和优化，这就需要多物理场、多尺度高性能计算计算的强耦合套件。核聚变数字孪生体的安全监管、不确定性量化和优化无疑是一项埃级大挑战。在这种情况下，拥有代用模型来提供快速估算和不确定性量化，对于探索和优化新的设计方案至关重要。但问题在于：精确的代用模型训练首先需要模拟数据。大量工作已经探索了求解器在环解决方案，以最大限度地训练这种代用模型。同样，也有人提出了创新方法，利用代型加速传统的高性能计算求解器。在这里，我们设计了一种新方法来实现这两点。通过将神经算子和高性能计算方法结合在一起，实现了一个自我完善的框架。随着越来越多的仿真在该框架内运行，代用程序会不断改进，而高性能计算仿真则会不断加速。在与核聚变相关的 MHD 模拟中，基于傅立叶神经算子的代理被用来为 Parareal（时间并行化）方法创建神经粗解器。Parareal 与大型 HPC 模拟特别相关，因为传统的空间并行化已经饱和，因此时间维度也需要并行化。这个神经-Parareal 框架是向利用 HPC 和 AI 融合迈出的一步，科学家和工程师可以从自动化、自我完善、越来越快的模拟中获益。这里开发的所有数据/代码均向社区开放。

{"title":"Neural-Parareal: Self-improving acceleration of fusion MHD simulations using time-parallelisation and neural operators","authors":"S.J.P. Pamela , N. Carey , J. Brandstetter , R. Akers , L. Zanisi , J. Buchanan , V. Gopakumar , M. Hoelzl , G. Huijsmans , K. Pentland , T. James , G. Antonucci , JOREK Team","doi":"10.1016/j.cpc.2024.109391","DOIUrl":"10.1016/j.cpc.2024.109391","url":null,"abstract":"<div><div>The fusion research facility ITER is currently being assembled to demonstrate that fusion can be used for industrial energy production, while several other programmes across the world are also moving forward, such as EU-DEMO, CFETR, SPARC and STEP. The high engineering complexity of a tokamak makes it an extremely challenging device to optimise, and test-based optimisation would be too slow and too costly. Instead, digital design and optimisation must be favoured, which requires strongly-coupled suites of multi-physics, multi-scale High-Performance Computing calculations. Safety regulation, uncertainty quantification, and optimisation of fusion digital twins is undoubtedly an Exascale grand challenge. In this context, having surrogate models to provide quick estimates with uncertainty quantification is essential to explore and optimise new design options. But there lies the dilemma: accurate surrogate training first requires simulation data. Extensive work has explored solver-in-the-loop solutions to maximise the training of such surrogates. Likewise, innovative methods have been proposed to accelerate conventional HPC solvers using surrogates. Here, a novel approach is designed to do both. By bootstrapping neural operators and HPC methods together, a self-improving framework is achieved. As more simulations are being run within the framework, the surrogate improves, while the HPC simulations get accelerated. This idea is demonstrated on fusion-relevant MHD simulations, where Fourier Neural Operator based surrogates are used to create neural coarse-solver for the Parareal (time-parallelisation) method. Parareal is particularly relevant for large HPC simulations where conventional spatial parallelisation has saturated, and the temporal dimension is thus parallelised as well. This Neural-Parareal framework is a step towards exploiting the convergence of HPC and AI, where scientists and engineers can benefit from automated, self-improving, ever faster simulations. All data/codes developed here are made available to the community.</div></div>","PeriodicalId":285,"journal":{"name":"Computer Physics Communications","volume":"307 ","pages":"Article 109391"},"PeriodicalIF":7.2,"publicationDate":"2024-10-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142446884","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"物理与天体物理","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0