We explore the formation of intermediate-mass black holes (IMBHs), potential seeds for supermassive black holes (SMBHs), via runaway stellar collisions for a wide range of star cluster (surface) densities (4 & times; 10(3) less than or similar to Sigma(h) less than or similar to 4 & times; 10(6) M-circle dot pc(-2)) and metallicities (0.01 less than or similar to Z less than or similar to 1.0 Z(circle dot)). Our sample of isolated ( > 1400) and hierarchical (30) simulations of young, massive star clusters with up to N = 1 . 8 & times; 106 stars includes collisional stellar dynamics, stellar evolution, and post-Newtonian equations of motion for black holes using the BIFRoST code. High stellar wind rates suppress IMBH formation at high metallicities ( Z greater than or similar to 0.2 Z(circle dot)), and low collision rates prevent their formation at low densities (Sigma(h) less than or similar to 3 & times; 10(4) M-circle dot pc(-2)). The assumptions about stellar wind loss rates strongly affect the maximum final IMBH masses ( M-center dot similar to 6000 M-circle dot versus M-center dot similar to 25 000 M-circle dot). The total stellar mass loss from collisions and collisionally boosted winds before t = 3 Myr can together reach up to 5-10 per cent of the final cluster mass. We present fitting formulae for IMBH masses as a function of host star cluster Sigma(h) and Z which can be used to seed SMBHs in high-resolution cosmological hydrodynamical simulations and in semi-analytic models for galaxy formation. Our results favour IMBH formation in dense low-metallicity environments similar to z similar to 10 JWST ( James Webb Space Telescope ) proto-globular clusters. IMBH formation is suppressed in the high-metallicity and low-density conditions of the local Universe.

ESPResSo++ is an open-source software package for molecular dynamics (MD) simulations with a particular emphasis on coarse-grained (CG) models of soft matter systems. Written in C++ with a flexible Python interface, it is designed for high-performance computing (HPC) environments and supports massively parallel simulations through MPI. The package enables simulations of polymers, membranes, colloids and complex fluids with a wide range of interaction models and advanced algorithms.

The GROMACS molecular dynamics simulation package has historically supported a number of molecular mechanics force fields, including Amber. Force fields have evolved over time and continue to do so through refinement of existing interactions and the definition of additional ones to more faithfully match real-world experiments and higher-level quantum chemistry calculations. Although users have provided a number of conversions of newer Amber force fields over the years, the two most recent Amber protein force fields, ff14SB and ff19SB, had not previously been officially ported to GROMACS. A technical challenge is that ff19SB introduces residue-specific backbone dihedral energy correction maps (CMAPs), which required new support in GROMACS for residue-aware force-field preprocessing and topology generation. Unlike older force fields such as CHARMM27, in which CMAPs are parameterized using only atom types, ff19SB extends this scheme with additional amino-acid typing. Here, we describe how we ported and validated the Amber ff14SB and ff19SB force fields for use with GROMACS as AMBER14SB and AMBER19SB, respectively. The ported force fields reproduce the Amber reference topologies exactly, with per-interaction energies matching to within the single-precision floating-point limit (< 10 −7 kJ/mol). To achieve this level of consistency between Amber and GROMACS, we resolved discrepancies in their handling of fundamental constants, parameter precision, and improper-dihedral atom ordering logic. In long unrestrained ubiquitin simulations, the converted ff19SB reproduces the conformational free-energy profile obtained with the native Amber implementation. This work provides the first official port of an Amber force field to GROMACS in more than a decade, improving feature parity between the two simulation packages. We provide a reproducible porting and validation procedure that can be used for modified variants and future versions of Amber force fields.

Sure-independence screening and sparsifying operator (SISSO) is an artificial intelligence (AI) method based on symbolic regression and compressed sensing widely used in materials science research. SISSO++ is its C++ implementation that employs MPI and OpenMP for parallelization, rendering it well-suited for high-performance computing (HPC) environments. As heterogeneous hardware becomes mainstream in the HPC and AI fields, we chose to port the SISSO++ code to GPUs using the Kokkos performance-portable library. Kokkos allows us to maintain a single codebase for both Nvidia and AMD GPUs, significantly reducing the maintenance effort. In this work, we summarize the necessary code changes we did to achieve hardware and performance portability. This is accompanied by performance benchmarks on Nvidia and AMD GPUs. We demonstrate the speedups obtained from using GPUs across the three most time-consuming parts of our code.

Electronic-structure theory is the foundation of the description of materials including multiscale modeling of their properties and functions. Obviously, without sufficient accuracy at the base, reliable predictions are unlikely at any level that follows. The software package FHI-aims has proven to be a game changer for accurate free-energy calculations because of its scalability, numerical precision, and its efficient handling of density functional theory (DFT) with hybrid functionals and van der Waals interactions. It treats molecules, clusters, and extended systems (solids and liquids) on an equal footing. Besides DFT, FHI-aims also includes quantum-chemistry methods, descriptions for excited states and vibrations, and calculations of various types of transport. Recent advancements address the integration of FHI-aims into an increasing number of workflows and various artificial intelligence (AI) methods. This Roadmap describes the state-of-the-art of FHI-aims and advancements that are currently ongoing or planned.

The high-performance computing (HPC) community has recently seen a substantial diversification of hardware platforms and their associated programming models. From traditional multicore processors to highly specialized accelerators, vendors and tool developers back up the relentless progress of those architectures. In the context of scientific programming, it is fundamental to consider performance portability frameworks, i.e., software tools that allow programmers to write code once and run it on different computer architectures without sacrificing performance. We report here on the benefits and challenges of performance portability using a field-line tracing simulation and a particle-in-cell code, two relevant applications in computational plasma physics with applications to magnetically-confined nuclear-fusion energy research. For these applications we report performance results obtained on four HPC platforms with server-class CPUs from Intel (Xeon) and AMD (EPYC), and high-end GPUs from Nvidia and AMD, including the latest Nvidia H100 GPU and the novel AMD Instinct MI300A APU. Our results show that both Kokkos and OpenMP are powerful tools to achieve performance portability and decent "out-of-the-box" performance, even for the very latest hardware platforms. For our applications, Kokkos provided performance portability to the broadest range of hardware architectures from different vendors.

ABSTRACT We benchmarked the performance of the GROMACS 2024 molecular dynamics (MD) code on a modern high‐performance computing (HPC) cluster with AMD CPUs on up to 65,536 CPU cores. We used five different MD systems, ranging in size from about 82,000 to 204 million atoms, and evaluated their performance using two different Message Passing Interface (MPI) libraries, Intel‐MPI and Open‐MPI. The largest system showed near‐perfect strong scaling up to 512 nodes or 65,536 cores, maintaining a parallel efficiency above 0.9 even at the highest level of parallelization. Energy efficiency for a given number of nodes was generally equal to or slightly better than parallel efficiency. We achieved peak performances of 687 ns/d for the 82k atom system, 116 ns/d for the 53M atom system, and about 35 ns/d for the largest 204M atom system. These results demonstrate that highly optimized software running on a state‐of‐the‐art HPC cluster provides sufficient computing power to simulate biomolecular systems at the mesoscale of viruses and organelles, and potentially small cells in the near future.

Science is and always has been based on data, but the terms "data-centric" and the "4th paradigm of" materials research indicate a radical change in how information is retrieved, handled and research is performed. It signifies a transformative shift towards managing vast data collections, digital repositories, and innovative data analytics methods. The integration of Artificial Intelligence (AI) and its subset Machine Learning (ML), has become pivotal in addressing all these challenges. This Roadmap on Data-Centric Materials Science explores fundamental concepts and methodologies, illustrating diverse applications in electronic-structure theory, soft matter theory, microstructure research, and experimental techniques like photoemission, atom probe tomography, and electron microscopy. While the roadmap delves into specific areas within the broad interdisciplinary field of materials science, the provided examples elucidate key concepts applicable to a wider range of topics. The discussed instances offer insights into addressing the multifaceted challenges encountered in contemporary materials research.

Hybrid density functional approximations (DFAs) offer compelling accuracy for ab initio electronic-structure simulations of molecules, nanosystems, and bulk materials, addressing some deficiencies of computationally cheaper, frequently used semilocal DFAs. However, the computational bottleneck of hybrid DFAs is the evaluation of the non-local exact exchange contribution, which is the limiting factor for the application of the method for large-scale simulations. In this work, we present a drastically optimized resolution-of-identity-based real-space implementation of the exact exchange evaluation for both non-periodic and periodic boundary conditions in the all-electron code FHI-aims, targeting high-performance CPU compute clusters. The introduction of several new refined Message Passing Interface (MPI) parallelization layers and shared memory arrays according to the MPI-3 standard were the key components of the optimization. We demonstrate significant improvements of memory and performance efficiency, scalability, and workload distribution, extending the reach of hybrid DFAs to simulation sizes beyond ten thousand atoms. As a necessary byproduct of this work, other code parts in FHI-aims have been optimized as well, e.g., the computation of the Hartree potential and the evaluation of the force and stress components. We benchmark the performance and scaling of the hybrid DFA based simulations for a broad range of chemical systems, including hybrid organic-inorganic perovskites, organic crystals and ice crystals with up to 30,576 atoms (101,920 electrons described by 244,608 basis functions).

Quantitative analysis of microstructural features on the nanoscale, including precipitates, local chemical orderings (LCOs) or structural defects (e.g. stacking faults) plays a pivotal role in understanding the mechanical and physical responses of engineering materials. Atom probe tomography (APT), known for its exceptional combination of chemical sensitivity and sub-nanometer resolution, primarily identifies microstructures through compositional segregations. However, this fails when there is no significant segregation, as can be the case for LCOs and stacking faults. Here, we introduce a 3D deep learning approach, AtomNet, designed to process APT point cloud data at the single-atom level for nanoscale microstructure extraction, simultaneously considering compositional and structural information. AtomNet is showcased in segmenting L12-type nanoprecipitates from the matrix in an AlLiMg alloy, irrespective of crystallographic orientations, which outperforms previous methods. AtomNet also allows for 3D imaging of L10-type LCOs in an AuCu alloy, a challenging task for conventional analysis due to their small size and subtle compositional differences. Finally, we demonstrate the use of AtomNet for revealing 2D stacking faults in a Co-based superalloy, without any stacking-faults-relevant samples in the training dataset, expanding the capabilities for automated exploration of hidden microstructures in APT data. AtomNet can thus recognize challenging microstructures, including nanoprecipitates with diameters above 2 nm, LCOs with diameters of about 1–2 nm without obvious compositional segregation, and even unforeseen planar defects by analyzing atom-atom environments. AtomNet pushes the boundaries of APT analysis, and holds promise in establishing precise quantitative microstructure-property relationships across a diverse range of metallic materials.

In solids, chemical short-range order (CSRO) refers to the self-organization of atoms of certain species occupying specific crystal sites. CSRO is increasingly being envisaged as a lever to tailor the mechanical and functional properties of materials. Yet quantitative relationships between properties and the morphology, number density, and atomic configurations of CSRO domains remain elusive. Herein, it is showcased how machine learning-enhanced atom probe tomography (APT) can mine the near-atomically resolved APT data and jointly exploit the technique's high elemental sensitivity to provide a 3D quantitative analysis of CSRO in a CoCrNi medium-entropy alloy. Multiple CSRO configurations are revealed, with their formation supported by state-of-the-art Monte-Carlo simulations. Quantitative analysis of these CSROs allows establishing relationships between processing parameters and physical properties. The unambiguous characterization of CSRO will help refine strategies for designing advanced materials by manipulating atomic-scale architectures.

With the end of Dennard scaling, specializing and distributing compute engines throughout the system is a promising technique to improve applications performance. For example, NVIDIA's BlueField Data Processing Unit (DPU) integrates programmable processing elements within the network and offers specialized network processing capabilities. These capabilities enable communication via offloads onto DPUs and present new application opportunities for offloading nonblocking or complex communication patterns such as collective communication operations. This paper discusses the lessons learned enabling DPU-based acceleration for collective communication algorithms by describing the impact of such offloaded collective operations on two applications: Octopus and P3DFFT++. We present new algorithms for the nonblocking MPI_Ialltoallv and blocking MPI_Allgatherv collective operations that leverage DPU offloading, which are used by the above applications, and evaluate them. Our experiments show a performance improvement in the range of 14% to 49% for P3DFFT++ and 17% for Octopus, even though the performance of those collectives in well-balanced OSU latency benchmarks shows comparable performance to well-optimized host-based implementations of these collectives. This demonstrates that taking into account load imbalance in communication algorithms can help improve application performance where such imbalance is common and large in magnitude.

Modern HPC systems are increasingly relying on greater core counts and wider vector registers. Thus, applications need to be adapted to fully utilize these hardware capabilities. One class of applications that can benefit from this increase in parallelism are molecular dynamics simulations. In this paper, we describe our efforts at modernizing the ESPResSo++ molecular dynamics simulation package by restructuring its particle data layout for efficient memory accesses and applying vectorization techniques to benefit the calculation of short-range non-bonded forces, which results in an overall three times speedup and serves as a baseline for further optimizations. We also implement fine-grained parallelism for multi-core CPUs through HPX, a C++ runtime system which uses lightweight threads and an asynchronous many-task approach to maximize concurrency. Our goal is to evaluate the performance of an HPX-based approach compared to the bulk-synchronous MPI-based implementation. This requires the introduction of an additional layer to the domain decomposition scheme that defines the task granularity. On spatially inhomogeneous systems, which impose a corresponding load-imbalance in traditional MPI-based approaches, we demonstrate that by choosing an optimal task size, the efficient work-stealing mechanisms of HPX can overcome the overhead of communication resulting in an overall 1.4 times speedup compared to the baseline MPI version.

Chemical short-range order (CSRO) refers to atoms of specific elements self-organising within a disordered crystalline matrix to form particular atomic neighbourhoods. CSRO is typically characterized indirectly, using volume-averaged or through projection microscopy techniques that fail to capture the three-dimensional atomistic architectures. Here, we present a machine-learning enhanced approach to break the inherent resolution limits of atom probe tomography enabling three-dimensional imaging of multiple CSROs. We showcase our approach by addressing a long-standing question encountered in body-centred-cubic Fe-Al alloys that see anomalous property changes upon heat treatment. We use it to evidence non-statistical B2-CSRO instead of the generally-expected D03-CSRO. We introduce quantitative correlations among annealing temperature, CSRO, and nano-hardness and electrical resistivity. Our approach is further validated on modified D03-CSRO detected in Fe-Ga. The proposed strategy can be generally employed to investigate short/medium/long-range ordering phenomena in different materials and help design future high-performance materials.

Thanks to the availability of huge amounts of data and improved computational resources, AI methods are gaining importance in scientific workflows, from image recognition and natural language processing to materials science. In many domains the usage of AI is under active investigation and first results show a tremendous potential, suggesting that AI will have significant impact way beyond the currently dominating examples of image and language processing.

For the first time, a real-time capable NBI code, which has a comparable fidelity to the much more computationally expensive Monte Carlo codes such as NUBEAM, has been coupled to the discharge control system of a tokamak. This implementation has been done at ASDEX Upgrade and is presented in this paper. Modifications to the numerical scheme of RABBIT for the time-dependent solution of the Fokker–Planck equation have been carried out to make it compatible with the non-equidistant time-steps, as they occur in real-time simulations. We demonstrate that this allows RABBIT to run in real-time both in a steady-state and time-dependent fashion and show and discuss an actual real-time simulation. Its accuracy is identified by comparing to offline RABBIT and TRANSP-NUBEAM runs (where more diagnostics are available for preciser inputs).

The effects of the electron inertia on the current sheets that are formed out of kinetic turbulence are relevant to understand the importance of coherent structures in turbulence and the nature of turbulence at the dissipation scales. We investigate this problem by carrying out 3D hybrid-kinetic Particle-in-Cell simulations of decaying kinetic turbulence with our CHIEF code. The main distinguishing feature of this code is an implementation of the electron inertia without approximations. Our simulation results show that the electron inertia plays an important role in regulating and limiting the largest values of current density in both real and wavenumber Fourier space, in particular, near and, unexpectedly, even above electron scales. In addition, the electric field associated with the electron inertia dominates most of the strongest current sheets. The electron inertia is thus important to accurately describe the properties of current sheets formed in turbulence at electron scales.

Computational methods and machine learning algorithms for automatic information extraction are crucial to enable data-driven materials science. These approaches are changing materials characterization and analytics, which often require a user-specified threshold to e.g. detect structure or symmetries in structures with defects. Here, we present a machine learning-based approach that directly works on the original periodic arrangements of atoms based on a three-dimensional convolutional neural network without any transformation of descriptors. Our approach shows a high classification accuracy and tolerance to the presence of random displacements and missing atoms. Experimentally, we successfully reconstruct the ordered L12 precipitates extracted from atom probe tomography data, consistent with segmentation based on isocomposition surfaces. The convolutional layers are essential for the simultaneous identification of compositional and structural information, which also give rise to its high tolerance. Our work advances machine learning-based crystal structure identification for incomplete crystal structural data.

Control over conductivity and carrier type (electrons and holes) defines semiconductors. A primary approach to target carrier concentrations involves introducing a small population of aliovalent impurity dopant atoms. In a combined synthetic and computational study, we assess impurity doping by introducing Bi and Sn into the prototype 2D Ruddlesden-Popper hybrid perovskite phenylethylammonium lead iodide (PEA_{2}PbI_{4}). Experimentally, we demonstrate that Bi and Sn can achieve n- and p-type doping, respectively, but the doping efficiency is low. Simulations show that Bi introduces a deep defect energy level (∼0.5 eV below the conduction band minimum) that contributes to the low doping efficiency, but, to reproduce the low doping efficiency observed experimentally, an acceptor level must also be present that limits n-type doping. Experiments find that Sn achieves p-dopant behavior and simulations suggest that this occurs through the additional oxidation of Sn defects. We also study how substitutional Bi incorporation can be controlled by tuning the electrochemical environment during synthesis. First-principles impurity doping simulations can be challenging; typical dopant concentrations constitute less than 0.01% of the atoms, necessitating large supercells, while a high level of theory is needed to capture the electronic levels. We demonstrate simulations of complex defect-containing unit cells that include up to 3383 atoms, employing spin-orbit coupled hybrid density functional theory. While p- and n-type behavior can be achieved with Sn and Bi, simulations and experiments provide concrete directions where future efforts must be focused to achieve higher doping efficiency.

Accelerated computing is becoming more diverse as new vendors and architectures come into play. Although platform-specific programming models promise ease of development and better control over performance, they still restrict the portability of scientific applications. As the OpenMP offloading specification becomes adopted by more compilers, this programming model stands out as a vendor-neutral portable approach to heterogeneous programming. In this study, we port a plasma physics oriented field line tracing code from a CPU-based MPI+OpenMP approach to a GPU accelerated version, using OpenMP’s offloading capabilities. We analyze GPU performance across different vendors with respect to the original CPU version and test both prescriptive and descriptive approaches to accelerator programming. A maximum 6× acceleration over the CPU implementation was achieved using OpenMP’s high-level offloading directives. In addition, we demonstrate portability across three different vendor GPUs with no code modifications.