SEARCH · Search NASA
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.
Cyber supply chain risk management (C-SCRM) programs must consider operations that depend on the lifecycles of digital components such as hardware, firmware, software, and services. We integrate academic literature, historical incidents, and existing standards to identify obstacles faced by C-SCRM programs.
Abstract Background Applying graph convolutional networks (GCN) to the classification of free-form natural language texts leveraged by graph-of-words features (TextGCN) was studied and confirmed to be an effective means of describing complex natural language texts. However, the text classification models based on the TextGCN possess weaknesses in terms of memory consumption and model dissemination and distribution. In this paper, we present a fast message passing network (FastMPN), implementing a GCN with message passing architecture that provides versatility and flexibility by allowing trainable node embedding and edge weights, helping the GCN model find the better solution. We applied the FastMPN model to the task of clinical information extraction from cancer pathology reports, extracting the following six properties: main site, subsite, laterality, histology, behavior, and grade. Results We evaluated the clinical task performance of the FastMPN models in terms of micro- and macro-averaged F1 scores. A comparison was performed with the multi-task convolutional neural network (MT-CNN) model. Results show that the FastMPN model is equivalent to or better than the MT-CNN. Conclusions Our implementation revealed that our FastMPN model, which is based on the PyTorch platform, can train a large corpus (667,290 training samples) with 202,373 unique words in less than 3 minutes per epoch using one NVIDIA V100 hardware accelerator. Our experiments demonstrated that using this implementation, the clinical task performance scores of information extraction related to tumors from cancer pathology reports were highly competitive.
The Electric Vehicle Secure Architecture Laboratory Demonstration (EV SALaD) program is a demonstration of cybersecurity best practices for high-power electric vehicle (EV) charging infrastructure led by Idaho National Laboratory (INL), in collaboration with other DOE National Laboratories participating in the EVs at Scale Consortium.a Sandia National Laboratories (SNL) and Pacific Northwest National Laboratory (PNNL) participated in the first 2-year (FY22-23) demonstration cycle for EV SALaD. This report documents the FY23 demonstration, the second in a series of demonstrations and collaborations in deploying and operating cybersecure EV charging infrastructure. It includes a summary of improvements from the FY22 demonstration, technical analysis of the FY23 demonstration, how the research demonstrates cyber-physical and cybersecurity best practices for high-power EV charging infrastructure, and related impacts to national and energy security. For EV SALaD, the FY22 demonstration focused on the detection, ranking, and prioritization of anomalous events for high-power EV charging. The FY23 demonstration additionally included the demonstration of cybersecurity best practices, which included protection and mitigation solutions to prevent, respond, and recover from anomalous events. During the demonstrations, the multi-lab EV SALaD team conducted a Test Effect Payload (TEP)b evaluation on extreme fast charger (XFC) hardware equipped with Cerberus, a detection and response solution, to demonstrate anomaly detection and mitigation cybersecurity best practices against cyber-enabled events.
While quantum annealers have emerged as versatile and controllable platforms for experimenting on correlated spin systems, the important phenomenology of magnetic memory and hysteresis remain unexplored on hardware designed to escape metastable states via quantum tunneling. Here, we present the first general protocol to experiment on magnetic hysteresis on programmable quantum annealers and implement it on three D-Wave superconducting qubit quantum annealers, using up to thousands of spins, for both ferromagnetic and disordered Ising models, and across different graph topologies. We observe hysteresis loops whose area depends nonmonotonically on quantum fluctuations, exhibiting both expected and unexpected features, such as disorder-induced steps and nonmonotonicities. Our work establishes quantum annealers as a platform for probing nonequilibrium emergent magnetic phenomena, thereby broadening the role of analog quantum computers into foundational questions in condensed matter physics.
HLS4ml (high level synthesis for machine learning) Is a Python package used to translate commonly used open-source machine learning models into HLS. This is useful in machine learning applications on FPGAs. Machine learning algorithms are only as fast as the hardware that they are used on, and some applications require high speed without sacrificing accuracy. In these situations, an FPGA is a good choice since it is faster than a CPU or a GPU, but programming an FPGA is difficult. This is where HLS4ml can be used to simplify the process, as a well-known learning model can be converted to HLS and more easily deployed onto an FPGA. There are many use cases for a machine learning algorithm running on an FPGA. For example, detectors in a particle accelerator cannot keep every event that they detect, and so a computer must decide which events to keep and which to discard. Using an FPGA with a machine learning algorithm would be a good way to keep as many events as possible.
hls4ml (high level synthesis for machine learning) Is a Python package used to translate commonly used open-source machine learning models into HLS. This is useful in machine learning applications on FPGAs. Machine learning algorithms are only as fast as the hardware that they are used on, and some applications require high speed without sacrificing accuracy. In these situations, an FPGA is a good choice since it is faster than a CPU or a GPU, but programming an FPGA is difficult. This is where hls4ml can be used to simplify the process, as a well-known learning model can be converted to HLS and more easily deployed onto an FPGA. There are many use cases for a machine learning algorithm running on an FPGA. For example, detectors in a particle accelerator cannot keep every event that they detect, and so a computer must decide which events to keep and which to discard. Using an FPGA with a machine learning algorithm would be a good way to keep as many events as possible.
Run to run variability in parallel programs caused by floating-point non-associativity has been known to significantly affect reproducibility in iterative algorithms, due to accumulating errors. Non-reproducibility can critically affect the efficiency and effectiveness of correctness testing for stochastic programs. Recently, the sensitivity of deep learning training and inference pipelines to floating-point non-associativity has been found to sometimes be extreme. It can prevent certification for commercial applications, accurate assessment of robustness and sensitivity, and bug detection. New approaches in scientific computing applications have coupled deep learning models with high-performance computing, leading to an aggravation of debugging and testing challenges. Here we perform an investigation of the statistical properties of floating-point non-associativity within modern parallel programming models, and analyze performance and productivity impacts of replacing atomic operations with deterministic alternatives on GPUs. We examine the recently-added deterministic options in PyTorch within the context of GPU deployment for deep learning, uncovering and quantifying the impacts of input parameters triggering run to run variability and reporting on the reliability and completeness of the documentation. Finally, we evaluate the strategy of exploiting automatic determinism that could be provided by deterministic hardware, using the Groq LPUTM accelerator for inference portions of the deep learning pipeline. We demonstrate the benefits that a hardware-based strategy can provide within reproducibility and correctness efforts.
Abstract We have developed a modular graphical user interface (GUI)-based program for use in genetic algorithm-based feedback-assisted wavefront shaping. The program uses a class-based structure to separate out the universal modules (e.g. GUI, multithreading, optimization algorithms) and hardware-specific modules (e.g. code for different SLMs and cameras). This modular design makes the program easily adaptable to a wide range of lab equipment, while providing easy access to a GUI, multithreading, and three optimization algorithms (phase-stepping, simple genetic, and microgenetic).
Simulating strongly coupled gauge theories at finite temperature and density is a longstanding challenge in nuclear and high-energy physics with fundamental implications for condensed matter physics. Here, we simulate such systems using minimally entangled typical thermal state (METTS) approaches, which combine classical random sampling with imaginary-time evolution, implementable on either classical or quantum computers, to estimate thermal averages of observables. We study 1+1-dimensional ${\mathbb{Z}}_{2}$ gauge theory coupled to spinless fermionic matter, which maps onto a local quantum spin chain. We benchmark both a classical matrix-product-state implementation of METTS and a recently proposed adaptive variational approach for near-term quantum devices, focusing on the equation of state and measures of fermion confinement. Of particular importance is the choice of basis for METTS sampling, which impacts both the sampling overhead and quantum circuit complexity. Our work sets the stage for future studies of strongly coupled gauge theories using classical and quantum hardware.
Training in Deep learning (DL) remains highly compute- and data-intensive, with I/O becoming a critical bottleneck as models and datasets scale. Recent studies report that data loading can dominate training time, especially on large-scale HPC systems with shared parallel file systems (PFS). Existing caching approaches either rely on single-tier designs or require intrusive modifications to training pipelines, limiting their portability and effectiveness. In this work, we present FitCache, a transparent drop-in framework for multi-tier caching to accelerate distributed DL training by coordinating fast local memory (e.g., DRAM, Persistent Memory (PMem)) and NVMe as hierarchical caches atop PFS. Our design adapts to hardware diversity, i.e., if NVMe is missing, memory transparently acts as a caching tier, ensuring stable performance. FitCache transparently intercepts I/O requests and issues concurrent fetches across all tiers, returning data from the fastest responder without centralized metadata or static redirection paths. FitCache adapts to dynamic workloads and heterogeneous clusters while maintaining POSIX compatibility. Experiments on Frontier (2048 GPUs) and smaller research clusters show that FitCache reduces training time by up to 40% and per-batch I/O latency by up to 71.6% compared to Lustre Orion PFS, offering a drop-in solution for scalable DL training.
Multiphysics codes links to a library called SINGE for the calculation of thermonuclear (TN) burn rates, but some current multiphysics codes do not attempt to leverage the support for parallel operation that SINGE provides. Our goal is to investigate implementations of the SINGE workflow and analyze how the use of a performance portability layer could reduce run time on CPU archi tectures while also supporting GPU architectures without requiring code modifications. We looked to the Kokkos C++ Performance Portability Ecosystem to implement hardware agnostic parallel patterns.
Defining a domain-specific language (DSL) that supports vector-calculus abstractions eases the porting of partial differential equation (PDE) solvers to specialized architectures. Sufficiently high-level abstractions empower users to express universal laws with sufficient generality that the laws must always hold true within their domain of validity. A broad class of PDE solvers employs stencil-based algorithms, the target domain of Berkeley Lab's stencil accelerator chip co-design project. First released as open-source in January 2026, the Formal software framework lays a foundation for defining an embedded DSL based on composable operators that implement mimetic numerical methods -- stencil algorithms that guarantee satisfaction of discrete versions of important vector calculus theorems. The Formal DSL will be the frontend to a new class of stencil-PDE accelerators developed jointly by LBNL, UHCL, and UC Berkeley through the DOE Competitive Portfolios for Computer Science Project. This offers the potential of an order of magnitude acceleration for this important category of computational methods to serve the DOE mission. Future work on the Formal DSL will facilitate software verification via type-safe templates that enable problem-specific correctness proofs relying upon generic function theory and carefully crafted unit tests.
Directly visualizing the trajectories of chemistry can unravel novel insights into the behavior of catalysts, gas phase reactions, photo-induced dynamics, and building blocks for quantum information processing. The ability of explicitly identifying, tracking, and tagging the exchange of matter, hence the annihilation and creation of new chemical species, can be best realized through a close coupling of theory and experiment. While the synchrotron-based characterization facilities propelled rapidly in its hardware, providing higher brightness, better resolution, and more precision, the software infrastructure is lagging. We developed DTCS (Digital Twin for Chemical Science) v.01, a central platform that faithfully mimics advanced instrumentations in Scientific User Facilities, by solving a variety of technical challenges in data acquisition, analysis, and model-driven interpretation. Rooted in physics and accelerated by AI, we validated this concept by direct comparison with precise experimental X-ray Photoelectron Spectroscopy (XPS) observations using a ubiquitous metal-water interfacial scenario, i.e., Ag/H2O as our main narrative. The DTCS v.01 input mirrors how the bench chemists work, with the output directly linked to the end station computer, thereby providing a user-friendly, knowledge-driven, and accessible user experience with mechanistic insights standardized in a way that are ready to be published, versioned, and transferred flexibly.
Modular Multi Agent Grid Emulations for Joined Breakdowns in Linked Generative Emulations (MMAREJBLIGE) introduces an agent-based modeling framework into real-time cyber-physical emulation to achieve a context-aware environment that introduces operator/attacker/external-condition variability to improve emulation fidelity and testing rigor. We detail our agent framework design, internal communication via message passing, and time synchronization, as well as the individual components of the system. We include a brief analysis of several scenarios run on a real-time, hardware-in-the-loop, Industrial Control Systems (ICS) test-bed which include normal operation, physical disruption, disruption with mitigation, and disruption with mitigation during a cyber denial-of-service (DOS) attack.
Large language models (LLMs) have shown remarkable potential in various domains but often lack the ability to access and reason over domain-specific knowledge and tools. In this article, we introduce Chemistry Agent Connecting Tool-Usage to Science (CACTUS), an LLM-based agent that integrates existing cheminformatics tools to enable accurate and advanced reasoning and problem-solving in chemistry and molecular discovery. We evaluate the performance of CACTUS using a diverse set of open-source LLMs, including Gemma-7b, Falcon-7b, MPT-7b, Llama3-8b, and Mistral-7b, on a benchmark of thousands of chemistry questions. Our results demonstrate that CACTUS significantly outperforms baseline LLMs, with the Gemma-7b, Mistral-7b, and Llama3-8b models achieving the highest accuracy regardless of the prompting strategy used. Moreover, we explore the impact of domain-specific prompting and hardware configurations on model performance, highlighting the importance of prompt engineering and the potential for deploying smaller models on consumer-grade hardware without a significant loss in accuracy. By combining the cognitive capabilities of open-source LLMs with widely used domain-specific tools provided by RDKit, CACTUS can assist researchers in tasks such as molecular property prediction, similarity searching, and drug-likeness assessment.
The Chemical Recommender System (CRS) is an open-source, high-performance toolkit that enables real-time similarity searches across the complete PubChem database (over 50 million molecules) using commodity hardware. The CRS addresses critical limitations in existing chemical informatics platforms through a novel vector database infrastructure, extensible model integration capabilities, and complete algorithmic transparency. The system implements a vector database deployment with partitioned indexing that achieves a ~60x speedup over traditional approaches. A containerized model integration framework allows researchers to seamlessly incorporate custom predictive models into the full-scale search and scoring pipeline, while complete configurability of search parameters, filtering logic, and scoring functions provides capabilities not available in existing black-box solutions. Beyond structural similarity, the CRS integrates OPERA QSAR models for thermophysical and toxicity predictions, RDKit synthetic accessibility scoring, and user-defined models to compute weighted final replacement scores. The complete system is accessible through an interactive web application supporting real-time progress monitoring, post-processing score re-weighting, automated PDF reporting, and batch processing capabilities.
Continuously parameterized two-qubit gates are a key feature of state-of-the-art trapped-ion quantum processors, as they have favorable error scalings and show distinct improvements in circuit performance over more restricted maximally entangling gatesets. In this work, we provide a comprehensive and pedagogical discussion on how to practically implement these continuously parameterized Mølmer–Sørensen gates on the Quantum Scientific Computing Open User Testbed, a low-level trapped-ion processor. To generate the arbitrary entangling angles, θ, we simply scale the amplitude of light used to generate the entanglement. However, doing so requires careful consideration of amplifier saturation as well as the variable light shifts that result. As such, we describe a method to calibrate and cancel the dominant fourth-order effects, followed by a dynamic virtual phase advance during the gate to cancel any residual light shifts, and find a linear scaling between θ and the residual light shift. Once we have considered and calibrated these effects, we demonstrate performance improvement with decreasing θ. Finally, we describe nuances of hardware control to transform the XX-type interaction of the arbitrary-angle Mølmer–Sørensen gate into a phase-agnostic and crosstalk-mitigating ZZ interaction.
All Hazards Analysis (AHA) is a framework developed by Idaho National Laboratory that provides capabilities to collect, store, analyze, and visualize critical infrastructure information. A core function of AHA is its ability to simulate faults or outages in networks of infrastructure originating from a plethora of causes, ranging from natural disasters to cyberattacks. AHA utilizes Hardware and Software Bills of Material (HBOM and SBOM, respectively) along with Known Exploited Vulnerabilities (KEVs) to document the potential attack vectors for each piece of infrastructure. The objective of this contribution to AHA was to create a visualization tool that could capture the small details held in each individual artifact as well as preserve the large-scale connections that link them together to aid threat modeling.