Search NASA⌕ Search

SEARCH · Search NASA

Results for “Stack”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 793 records · Page 44
All sourcesNASA NTRSDOE OSTI
DOE OSTI2025

Surface acoustic waves and lattice vibrations in two-dimensional Ti 3 ⁢C 2 ⁢𝑇 𝑥 (𝑇 =O, F) MXene films

We investigated surface acoustic wave (SAW) propagation and lattice vibrations in two-dimensional (2D) titanium carbide (Ti 3 ⁢C 2 ⁢𝑇 x ) MXene films as a function of surface termination and layer stacking, using atomistic simulations. We found that SAW propagation velocity is highly sensitive to both single-layer properties and interlayer bonding. Surface terminations significantly modulate wave behavior, with oxygen and fluorine terminations producing distinct effects on wave propagation, with oxygen-terminated monolayers exhibiting 20% higher wave speeds than fluorine counterparts due to strengthened intralayer bonds. Key observations include the transition from one to two layers causing wave speed variations, and the development of interlayer modes that generate more dispersed lattice vibrations. As the film layer thickness increases, SAW propagation becomes predominantly confined to the upper surface, with coherence of vibrational modes diminishing in multilayer structures. These findings suggest MXene terminations and layer stacking are crucial parameters for controlling SAW behavior, offering promising avenues for novel acoustic wave device applications.

2-dimensional systems↗
DOE OSTI2024

Photonuclear cross sections for 197 Au: An update on the gold standard

Cross sections for the 197 Au(γ, n) reaction are broadly used in nuclear physics as a standard for normalizing photonuclear reaction cross-section data at photon beam energies above approximately 8 MeV. In this paper, we report cross-section measurements for the 197 Au(γ, n) 196 Au g+m1 reaction at beam energies from 13 to 31 MeV. Our measurements provide the first cross-section data for this reaction at beam energies above 20 MeV, enabling the use of this reaction as a cross-section standard up to 30 MeV. Also, this work provides first cross-section measurements for the 197 Au(γ, n) 196 Au m2 reaction. In addition, we measured cross-section data for the 197 Au(γ, 3n) 194 Au reaction, which can be used as a cross-section standard above about 25 MeV. These measurements were performed using a new target activation method that is based on the angle-energy correlation of the laser Compton- scattered photon beams at the High Intensity Gamma-ray Source (HIγS). The technique enables measuring photonuclear reaction cross-sections at several discrete beam energies concurrently via a single irradiation on a stack of different targets. Measurements were carried out by irradiating a stack of concentric-ring targets consisting of Au, TiO 2 , Zn, Os, and Au (in order of the γ-ray beam direction). Our data for the 197 Au(γ,n) 196 Au g+m1 reaction in the energy range of 13 to 20 MeV are in good agreement with existing ones measured using monoenergetic γ-ray beams, but differ from data acquired using a bremsstrahlung γ-ray beam. Also, above 18 MeV, our data for the 197 Au(γ,n) 196 Au g+m1 and 197 Au(γ,n) 196 Au m2 reactions differ significantly from the most recent TENDL and JENDL evaluations, suggesting a need to update these data libraries. Furthermore, the TENDL evaluation and existing data are consistent with our data for the 197 Au(γ,3n) reaction, but differ significantly from the JENDL evaluation above 26 MeV.

190 ≤ A ≤ 219↗
DOE OSTI2025

Reaching the prolate-oblate boundary at 𝑁=116 via first fragmentation of a 198 Pt beam: Sharp transition to triaxiality in 189 Ta

High-spin isomers in very-neutron-rich 𝐴≈190 Hf-Ta-W nuclei were populated via the pioneering fragmentation of a 198 Pt primary beam at the National Superconducting Cyclotron Laboratory. The nuclei were implanted in a Si detector stack surrounded by the Gamma-Ray Energy Tracking In-beam Nuclear Array (GRETINA) to detect delayed 𝛾 rays, providing first level schemes using 𝛾−𝛾 coincidence data from isomeric decays in this previously inaccessible region of the nuclear chart. Here, a sudden transition to a strong triaxial shape is observed in the very-neutron-rich 189 Ta (𝑁 = 116) nucleus from axially prolate shapes in lighter Ta isotopes, providing a critical experimental benchmark for competing theoretical predictions of nuclear-shape evolution.

150 ≤ A ≤ 189↗
DOE OSTI2025

Tracing cosmic gas in filaments and halos: Low-redshift insights from the kinematic Sunyaev-Zel’dovich effect

In this work, we leverage cosmic microwave background (CMB) data from the Atacama Cosmology Telescope (ACT) and LSS data from the imaging survey conducted by the Dark Energy Spectroscopic Instrument (DESI) to study the distribution of gas around galaxy groups at low redshift, z ≈ 0.3, via the kinematic Sunyaev-Zel'dovich (kSZ) effect. In particular, we perform velocity-weighted stacking on the photometric galaxies from the Bright Galaxy Survey (BGS) to isolate the monopole and quadrupole of the kSZ signal, orienting the stacked images along 2D filaments identified using the Hessian of the projected gravitational potential. We find a 7.2σ detection in the monopole of the signal (i.e., the gas density profile) and a 4σ detection in the quadrupole (m = 2), constituting the first measurement of the alignment between gas distribution and the cosmic web through the kSZ effect. As it is a linear probe of the local gas density, the kSZ has heightened sensitivity to the warm-hot intergalactic medium (WHIM), which is believed to house the majority of the "missing baryons."Mapping out the gas density at low redshifts, as enabled by our measurements, is crucial for weak lensing surveys, for which the impact of baryons on small scales is a major impediment. We compare the anisotropic signal against two hydrodynamical simulations, TNG300-1 and Illustris, which have very different baryonic feedback prescriptions. We find that the anisotropic signal measured in the data is comparable but slightly larger and more extended compared with the simulations. Further, this suggests that there is excess accretion and feedback taking place through the filaments, hinting at the possible presence of spin-filament alignment of the BGS objects.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2024

Evolution of structure and spectroscopic properties of a new 1,3-diacetylpyrene polymorph with temperature and pressure

A new polymorph of 1,3-diacetylpyrene has been obtained from its melt and thoroughly characterized using single-crystal X-ray diffraction, steady-state UV–Vis spectroscopy and periodic density functional theory calculations. Experimental studies covered the temperature range from 90 to 390 K and the pressure range from atmospheric to 4.08 GPa. Optimal sample placement in a diamond anvil cell according to our previously presented methodology ensured over 80% data coverage up to 0.8 Å for a monoclinic sample. Unrestrained Hirshfeld atom refinement of the high-pressure crystal structures was successful and anharmonic behavior of carbonyl oxygen atoms was observed. Unlike the previously characterized polymorph, the structure of 2°AP-β is based on infinite π-stacks of antiparallel 2°AP molecules. 2°AP-β displays piezochromism and piezofluorochromism which are directly related to the variation in interplanar distances within the π-stacking. The importance of weak intermolecular interactions is reflected in the substantial negative thermal expansion coefficient of -55.8 (57) MK -1 in the direction of C—H∙∙∙O interactions.

36 MATERIALS SCIENCE↗
DOE OSTI2025

Real-time Simulation Model of a Vanadium Redox Flow Battery Energy Storage System with Power Electronics Integration

This paper presents a real-time (RT) model of a 20 kW multi-stack vanadium redox flow battery (VRFB) system including power electronics converters in a single real-time framework. Implemented on a Typhoon HIL 604 device, this model is designed to simulate the operation of the flow battery with system-level controls. To provide a realistic overview of a VRFB system, the model incorporates a closed-loop control system for managing the speeds of two centrifugal pumps and an electro-thermal model of the battery stacks, enabling realtime temperature estimation. Furthermore, a two-stage power electronics topology and associated control solution are presented, operating with distinct strategies for interfacing the battery model with the main grid and a local load. Results from simulations demonstrate that the VRFB model is well suited for RT environment, and system can manage power during both charging and discharging cycles, while a SCADA monitor is used to display critical variables for safe, efficient, and reliable operation of the VRFB systems.

Rezende Da Costa Reis Kimpara, Renata [ORNL] (ORCI↗
DOE OSTI2024

TunIO: An AI-powered Framework for Optimizing HPC I/O

I/O operations are a known performance bottleneck of HPC applications. To achieve good performance, users often employ an iterative multistage tuning process to find an optimal I/O stack configuration. However, an I/O stack contains multiple layers, such as high-level I/O libraries, I/O middleware, and parallel file systems, and each layer has many parameters. These parameters and layers are entangled and influenced by each other. The tuning process is time-consuming and complex. In this work, we present TunIO, an AI-powered I/O tuning framework that implements several techniques to balance the tuning cost and performance gain, including tuning the high-impact parameters first. Furthermore, TunIO analyzes the application source code to extract its I/O kernel while retaining all statements necessary to perform I/O. It utilizes a smart selection of high-impact configuration parameters of the given tuning objective. Finally, it uses a novel Reinforcement Learning (RL)-driven early stopping mechanism to balance the cost and performance gain. Experimental results show that TunIO leads to a reduction of up to ≈73% in tuning time while achieving the same performance gain when compared to H5Tuner. It achieves a significant performance gain/cost of 208.4 MBps/min (I/O bandwidth for each minute spent in tuning) over existing approaches under our testing.

Rajesh, Neeraj↗
DOE OSTI2025

Understanding Reliability Trade-Offs in 1T-nC and 2T-nC FeRAM Designs

Ferroelectric random access memory (FeRAM) is a promising candidate for energy-efficient nonvolatile memory, particularly for logic-in-memory and compute-in-memory (CIM) applications. Among the available cell architectures, One-Transistor–n-Capacitor (1T-nC) and two-transistor–n-capacitor (2T-nC) FeRAMs each offer distinct trade-offs in density, scalability, and reliability. In this work, we present a comparative study of these two architectures under both dimensional scaling ( XY/Z shrinkage) and vertical integration (increasing stacked capacitors per cell). Using technology computer-aided design (TCAD) and circuit-level simulations, we analyze how scaling impacts ferroelectric capacitance, parasitic coupling, and floating-node (FN) dynamics, which together dictate sense margin (SM) and read stability. A key mitigation strategy—floating unselected capacitors—is applied to both architectures, effectively decoupling the SM from the number of stacked capacitors and enabling tractable analysis across scaling regimes. Results show that 1T-nC suffers more from charge sharing with the bitline (BL), while 2T-nC benefits from transistor isolation and stronger low-voltage sensing at the cost of increased area. By systematically evaluating these behaviors across scaling directions, this work establishes the reliability trade-offs of 1T-nC and 2T-nC cells and provides design guidelines for high-density, vertically integrated FeRAM systems.

1T-nC↗
DOE OSTI2024

Mitigating the Effects of Au-Al Intermetallic Compounds Due to High-Temperature Processing of Surface-Electrode Ion Traps

Stringent physical requirements need to be met for the high-performing surface-electrode ion traps used in quantum computing and timekeeping. In particular, these traps must survive a high-temperature environment for vacuum chamber preparation and support high RF voltage on closely spaced electrodes. Due to the use of gold wire bonds on aluminum pads, intermetallic growth can lead to wire bond failure via breakage or high resistance, limiting the lifetime of a trap assembly to a single multiday bake at 200 ° C. Using traditional thick metal stacks to prevent intermetallic growth, however, can result in trap failure due to RF breakdown events. Through high-temperature experiments, we conclude that an ideal metal stack for ion traps is Ti/Pt/Au (20/100/250 nm), which allows for a cumulative bakeable time of roughly 86 days without compromising the trap voltage performance. This increase in the bakeable lifetime of ion traps will remove the need to discard otherwise functional ion traps when vacuum hardware is upgraded, which will greatly benefit ion trap experiments.

Haltli, Raymond A.↗
DOE OSTI2025

Demonstration of Vertical 2T-nC FeRAM Hybrid Cell and Its Scalability for High-Density 3-D Ferroelectric Capacitor Memory

In this work, we present a comprehensive experimental and modeling study on the scaling of vertical 2T-nC ferroelectric random access memory (FeRAM) hybrid cells, comprising n metal-ferroelectric–metal (MFM) capacitors, to demonstrate a high-performance and high-density 3-D capacitor memory. Our contributions include: 1) successful process integration of vertical 2T-3C FeRAM cells by stacking MFM structures on top of Si CMOS transistors; 2) experimental validation of memory cell functionality, confirming the feasibility of the vertical 2T-nC FeRAM architecture; 3) an analysis of scaling effects on parasitic capacitance in densely integrated 3-D arrays, using 3-D technology computer-aided design (TCAD) simulations; 4) exploration of aggressive stacking of write bitlines (WBLs) to enhance memory density, where ferroelectric linear capacitance ( C FE ) enables self-boosted inhibition under the V W /2 scheme, but renders the V W /3 scheme ineffective due to intolerable write disturbances; and 5) assessment of horizontal scaling, revealing significant increases in read disturbances caused by interplane capacitance between adjacent WBLs ( C Z ). This work represents an early exploration into the potential of 2T-nC FeRAM as a scalable and efficient 3-D memory solution.

42 ENGINEERING↗
DOE OSTI2024

Unveiling and Mapping Polymorphs in Fluorite Y2TiO5 Using 4D-STEM and Unsupervised Machine Learning

Y2TiO5 belongs to the Ln2TiO5 (Ln = lanthanide or Y) family of ceramic materials and exhibits a range of desirable material properties such as radiation tolerance, frustrated magnetism, and large dielectric constant. However, understanding the complex crystal structure of Y2TiO5 remains elusive, given that Y2TiO5 can adopt multiple polymorphs such as cubic, orthorhombic, and hexagonal phases within the lattice. In this work, we report a detailed structural analysis of Y2TiO5 using four-dimensional scanning transmission electron microscopy coupled with unsupervised machine learning. The pyrochlore nanodomains, characterized by the ordered arrangement of yttrium cations on the A site of their A2BO5 structure, are present within the matrix of a predominantly fluorite-structured Y2TiO5 along with a third polymorph, the hexagonal phase. The pyrochlore phase is found to form 2 nm boundary regions around hexagonal phase stacking faults, highlighting the potential influence of the hexagonal phase on the occurrence and distribution of the pyrochlore phase. Lastly, we identify a unique pyrochlore phase with asymmetric arrangement of cation ordering along a single planar direction. Our findings provide invaluable insights into the possible mechanisms stabilizing pyrochlore nanodomains within the fluorite lattice of Y2TiO5.

36 MATERIALS SCIENCE↗
DOE OSTI2026

How state transitions balance photosynthetic electron transport in plants – a quantitative study

In plants, the process of state transition regulates the allocation of sunlight energy between Photosystem II (PSII) and PSI. However, the implications of state transitions for harmonizing electron transport rates between photosystems, and a full quantitative picture of this process, remain underexplored. We integrated quantitative biology (biochemical and biophysical approaches) with in vivo spectroscopy on wild-type Arabidopsis and protein phosphorylation mutants. This combination facilitated monitoring of Chl redistribution and its functional implications for light harvesting and electron transport. Our findings demonstrate the reallocation of 12% of highly phosphorylated ‘extra’ light-harvesting complex II under state 2 from stacked to unstacked thylakoids. This reduces the number of Chls per PSII from 216 to 182, while increasing the number in PSI from 187 to 223. Such Chl redistribution compensates for differences in photosystem stoichiometry and photochemical quantum efficiencies, thereby precisely synchronizing electron transport rates in both photosystems. Mutant analyses corroborate that this regulatory mechanism involves reversible phosphorylation. We inferred that state transitions optimize linear electron transport, leaving no additional capacity for cyclic electron transport. Furthermore, the results suggest that the controversies about long-range migration of LHCII from stacked to unstacked thylakoid domains arise from differences in phosphorylation levels.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2024

Exploring Architectural-Aware Affinity Policies in Modern HPC Runtimes

Modern commodity and High-Performance Computing (HPC) systems are evolving with complex CPU architectures. These architectures now feature higher core and NUMA domain counts and implement features such as hyperthreading. When considering significant differences in hardware configurations, library availability, and hardware-tailored system/software stacks, which could substantially vary from one system to another, performance portability is hard to achieve. Throughout the years, this trend resulted in an increasingly high burden on application developers to fine-tune their workloads for each architecture. This work explores how hardware-dependent aspects such as locality/process/thread affinity affect performance in modern CPU architectures. We focus our study on the Global Memory and Threading (GMT) distributed runtime system as a representative of Partitioned Global Address Space (PGAS) software stacks commonly adopted for productivity. In particular, to appreciate performance implications, we evaluate GMT’s thread affinity policies, and, introduce two new ones which exploit architectural awareness. Finally, we explore alternative NUMA configurations via different process bindings and perform a scalability study on three HPC clusters with varying CPU architectures and NUMA layouts. Our analysis indicates that more complex architectures are more affected by affinity and binding policies and highlights the importance of setting proper runtime configurations to achieve superior performance.

Di Dio Lavore, Ian↗
DOE OSTI2025

ChatHPC: Building the Foundations for a Productive and Trustworthy AI-Assisted HPC Ecosystem

ChatHPC democratizes large language models for the high-performance computing (HPC) community by providing the infrastructure, ecosystem, and knowledge needed to apply modern generative AI technologies to rapidly create specific capabilities for critical HPC components while using relatively modest computational resources. Our divide-and-conquer approach focuses on creating a collection of reliable, highly specialized, and optimized AI assistants for HPC based on the cost-effective and fast Code Llama fine-tuning processes and expert supervision. We target major components of the HPC software stack, including programming models, runtimes, I/O, tooling, and math libraries. Thanks to AI, ChatHPC provides a more productive HPC ecosystem by boosting important tasks related to portability, parallelization, optimization, scalability, and instrumentation, among others. With relatively small datasets (on the order of KB), the AI assistants, which are created in a few minutes by using one node with two NVIDIA H100 GPUs and the ChatHPC library, can create new capabilities with Meta’s 7-billion parameter Code Llama base model to produce high-quality software with a level of trustworthiness of up to 90% higher than the 1.8-trillion parameter OpenAI ChatGPT-4o model for critical programming tasks in the HPC software stack.

Young, Aaron [ORNL] (ORCID:0000000254484667)↗
DOE OSTI2025

Degradation of Fuel Cell Membrane Electrode Assemblies from Buses Operated More than 25,000 h

This study investigates the performance losses and degradation of proton-exchange-membrane fuel-cell stacks taken from the Alameda Contra Costa Transit District (AC Transit) bus system (Alameda and Contra Costa counties, California, United States) that were operated for over 25,000 h. Here, we focus on the origin of differences in electrochemical performance between beginning-of-life (BOL) and end-of-life states as well as diagnostic data acquired during the lifetime of the buses. In doing so, we employ in- and ex- situ characterization methods such as polarization curves, electrochemical impedance spectroscopy, electron microscopy, and X-ray characterization. Uniform degradation of the catalyst layer including Pt agglomeration/migration and electrode thinning was observed in all of the post-teardown measurements compared to BOL materials resulting from years of field operation. Despite these changes, the measured post-teardown performance suggests a sufficient output for the expected load, which indicate factors other than degradation of the membrane-electrode assemblies (MEAs) are likely responsible for the decommissioning of the stacks. The findings indicate that these MEA materials can enable long lifetime in fuel-cell vehicles, if the MEAs are not subjected to adverse operating conditions. The results also highlight the need for more in-vehicle diagnostics to maximize the lifetime of fuel cell vehicle (FCV) powerplants.

25 ENERGY STORAGE↗
DOE OSTI2025

Efficient Xml Interchange (exi) For Python (expy)

EXPy provides a native Python interface into the LF Energy EVerest V2G protocol stack. The protocol stack is implemented in C/C++ and compiled into shared object libraries. EXPy provides the Python Ctypes translation of the C/C++ libraries for use with pure Python software. This project eliminates the need for integrating Python with third-party communications applications and greatly reduces the code base and improves performance. The other major benefit is the ability for EXPy to support new EXI based protocols as additional V2G standards are produced (e.g. upgrade from ISO 15118-2 to ISO 15118-20).

Rohde, Kenneth [Idaho National Laboratory (INL), I↗
DOE OSTI2026

matsim-agents v1.0

matsim-agents is a multi-agent AI framework for atomistic materials simulation and discovery. It orchestrates large language models (LLMs), machine-learned interatomic potentials (MLIPs), and DFT codes into a single agentic loop running on laptops and DOE leadership-class supercomputers. MULTI-AGENT ORCHESTRATION A LangGraph state machine with three nodes: a Planner that converts a natural-language research objective into structured tasks; an Executor that dispatches atomistic tools and loops until the queue is empty; and an Analyst that summarizes results into a human-readable report. State is checkpointed after every step and human-in-the-loop gates can be inserted at any edge. HYPOTHESIS-DRIVEN DISCOVERY CHAT An interactive REPL (matsim-agents chat) that couples LLM dialogue with atomistic simulation. Chemical formulas are automatically detected in conversation turns and trigger a full crystal-phase exploration: structure generation → relaxation → stability scoring → result injection back into the conversation, creating a closed hypothesis-refinement loop. CRYSTAL PHASE ENUMERATION Given a composition, the phase explorer enumerates prototypes by stoichiometry: elemental (fcc/bcc/hcp/sc/diamond), binary 1:1 (rocksalt/CsCl/zincblende/ wurtzite/fluorite/rutile), ternary 1:1:3 (cubic perovskite), ternary 1:2:4 (perovskite + spinel), quaternary 1:1:2:6 (Fm-3m double perovskite). 2-D prototypes (graphene, h-BN, MoS2 2H/1T) and multilayer stacking are also supported via --include-2d and --num-layers. SUPERCELL GENERATION AND SITE DECORATION Auto-tiling to a minimum atom count (--min-atoms), explicit NxNxN tiling (--supercell), symmetry-distinct site decorations (--n-orderings), and isotropic lattice-scale sweeps (--lattice-scales) for volume bracketing. MLFF RELAXATION AND STABILITY SCORING HydraGNN (multi-headed GNN) drives structure relaxation via ASE with FIRE, BFGS, or BFGSLineSearch. Stability output: delta-E/atom ranking across phases and a max-residual-force dynamical-stability proxy. Other MLIPs (MACE, NequIP, Orb) can be plugged in through the same interface. DFT BACKENDS Quantum ESPRESSO pw.x and VASP 6.6 are first-class labellers. Both have validated GPU builds and SLURM/PBS launchers for three DOE platforms: Frontier (AMD MI250X, ROCm), Aurora (Intel PVC, oneAPI), Perlmutter (NVIDIA A100, CUDA). QE produces ~100 binaries (pw.x, ph.x, epw.x, ...). VASP supports scf, relax, vc-relax, and vc-relax-shape run types. ACTIVE-LEARNING LOOP matsim-agents al run CONFIG.yaml drives an iterative HydraGNN-DFT loop: MD generates candidates → ensemble/MC-dropout uncertainty selects the most informative → DFT labels them in parallel inside one allocation → dataset grows → HydraGNN retrains → repeat. DFT backend is a single YAML toggle (dft.backend: vasp | qe). LLM-generated seed structures are supported (no curated POSCAR library needed). Config uses ${VAR}, ${VAR:-default}, ${VAR:?msg} shell-style substitution for cross-user/cross-site portability. LLM BACKENDS Ollama (local, default), vLLM (HPC multi-GPU serving), OpenAI, Anthropic, HuggingFace Transformers+Accelerate. Selected at runtime via flag or env var with no code changes. HPC PORTABILITY Same Python entry points run on Frontier (ROCm 7.2), Aurora (oneAPI), and Perlmutter (CUDA 12). DFT and ML stacks are never co-loaded in the same shell; they couple through the scheduler and filesystem. Advanced multi-node launchers (serve, discovery-chat, single-relaxation, active-learning, QE warm-start) are provided for all three platforms. CODABENCH COMPETITION BUNDLE A self-contained benchmark: 159 atomistic test structures across 11 material classes, 5 tasks (formation energy, forces, ML relaxation, AI-DFT relaxation, phase stability ranking), public/private leaderboard split (30/70), and four ready-to-run baselines: MACE-MP-0, HydraGNN, UMA, AllScAIP.

Lupo Pasini, Massimiliano [Oak Ridge National Labo↗
DOE OSTI2024

MAGMA: Enabling exascale performance with accelerated BLAS and LAPACK for diverse GPU architectures

MAGMA (Matrix Algebra for GPU and Multicore Architectures) is a pivotal open-source library in the landscape of GPU-enabled dense and sparse linear algebra computations. With a repertoire of approximately 750 numerical routines across four precisions, MAGMA is deeply ingrained in the DOE software stack, playing a crucial role in high-performance computing. Notable projects such as ExaConstit, HiOP, MARBL, and STRUMPACK, among others, directly harness the capabilities of MAGMA. In addition, the MAGMA development team has been acknowledged multiple times for contributing to the vendors’ numerical software stacks. Looking back over the time of the Exascale Computing Project (ECP), we highlight how MAGMA has adapted to recent changes in modern HPC systems, especially the growing gap between CPU and GPU compute capabilities, as well as the introduction of low precision arithmetic in modern GPUs. We also describe MAGMA’s direct impact on several ECP projects. Maintaining portable performance across NVIDIA and AMD GPUs, and with current efforts toward supporting Intel GPUs, MAGMA ensures its adaptability and relevance in the ever-evolving landscape of GPU architectures.

97 MATHEMATICS AND COMPUTING↗