Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “edge computer”

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 1,099 records · Page 61
All sourcesNASA NTRSDOE OSTI
DOE OSTI2020

A Scalable Parallel Hypergraph Generator (HyGen)

Graphs are extensively used to model real-world complex systems. An edge in a graph can model pairwise relationships. However, multiway relationships (connections between three or more vertices) are common in many complex systems such as cellular process, image segmentation, and circuit design. A graph edge cannot model multiway relationships. A hypergraph, which can connect more than two vertices, is thus a better option to model multiway relationships. A large-scale hypergraph analysis has the potential to find useful insights from a complex system and assist in knowledge discovery. Currently a limited number of hypergraphs exists that are representative of real-world datasets. Moreover, real-world hypergraph datasets are small in size and inadequate to incorporate future needs. A graph generator that can produce large-scale synthetic hypergraphs can solve the above mentioned problems. In this paper, we present a scalable parallel hypergraph generator (HyGen) based on the Message Passing Interface (MPI) standard. To generate hypergraphs, HyGen takes the following parameter values as inputs: i) number of vertices, ii) number of hyperedges, iii) number of clusters, iv) vertex distribution, v) hyperedge distribution, vi) local cluster cardinality, and vii) global cluster cardinality. We have demonstrated that HyGen can generate hypergraphs of various sizes in a scalable fashion. HyGen takes approximately four minutes to generate a hypergraph with 4.8 million vertices, 1.6 million hyperedges, and 800 clusters using 1,024 processes on a leadership class computing platform. Our strong and weak scaling experiments on supercomputers demonstrate that HyGen can quickly create large-scale hypergraphs in a parallel manner, thus providing a useful capability for hypergraph analysis.

Hasan, S M Shamimul↗
DOE OSTI2023

Laboratory Directed Research and Development Program: FY 2021 Completed Projects

Oak Ridge National Laboratory (ORNL) is the US Department of Energy’s (DOE’s) largest multiprogram science, technology, and energy laboratory. It possesses distinctive capabilities in neutron science, computing, advanced materials, nuclear science and technology, and other fields. Using these capabilities, ORNL conducts basic and applied R&D to support DOE’s overarching mission “to ensure America’s security and prosperity by addressing its energy, environmental and nuclear challenges through transformative science and technology solutions.”1 As a national resource, ORNL also applies its capabilities and skills to specific needs of other federal agencies and customers through the DOE Strategic Partnership Projects (SPP) Program. Information about the laboratory and its programs is available on the ORNL website.2 The Laboratory Directed Research and Development (LDRD) Program at ORNL operates under the authority of the DOE Order 413.2C, “Laboratory Directed Research and Development,”3 which establishes DOE’s requirements for the program while providing the laboratory director broad flexibility for program implementation. The LDRD Program funds are obtained through a charge to all laboratory programs. Although it represents a relatively small portion of the overall research budget, the LDRD Program plays an essential role in maintaining the laboratory’s ability to respond to national needs. The program allows ORNL to improve its distinctive capabilities and to enhance its ability to conduct cutting-edge R&D. In accordance with the DOE order, R&D projects funded through the LDRD Program at ORNL support the following goals: (1) maintain the scientific and technical vitality of the laboratory; (2) enhance the laboratory’s ability to address future DOE missions; (3) foster creativity and stimulate exploration of forefront areas of science and technology; (4) serve as a proving ground for new concepts in R&D; and (5) support high-risk, potentially high-value R&D. This report provides an overview of the LDRD Program at ORNL in FY 2021 and contains summaries of all LDRD research projects that concluded between October 1, 2020, and September 30, 2021.

99 GENERAL AND MISCELLANEOUS↗
DOE OSTI2024

Laboratory Directed Research and Development Program: FY 2023 Completed Projects Report

Oak Ridge National Laboratory (ORNL) is the US Department of Energy’s (DOE’s) largest multiprogram science, technology, and energy laboratory. It possesses distinctive capabilities in a variety of fields, such as neutron science, computing, advanced materials, and nuclear science and technology. Using these capabilities, ORNL conducts basic and applied research and development (R&D) to support DOE’s overarching mission “to ensure America’s security and prosperity by addressing its energy, environmental and nuclear challenges through transformative science and technology solutions.” As a national resource, ORNL also applies its capabilities and skills to the specific needs of other federal agencies and customers through the DOE Strategic Partnership Projects (SPP) Program. Information about the laboratory and its programs is available on the ORNL website. The Laboratory Directed Research and Development (LDRD) Program at ORNL operates under the authority of the DOE Order 413.2C, Laboratory Directed Research and Development, which establishes DOE’s requirements for the program while providing the laboratory director broad flexibility for program implementation. The LDRD Program funds are obtained through a charge to all laboratory programs. Although it represents a relatively small portion of the overall research budget, the LDRD Program plays an essential role in maintaining the laboratory’s ability to respond to national needs. The program allows ORNL to improve its distinctive capabilities and to enhance its ability to conduct cutting-edge R&D. This report provides an overview of the LDRD Program at ORNL in FY 2023 and contains summaries of all the LDRD research projects that concluded between October 1, 2022, and September 30, 2023.

99 GENERAL AND MISCELLANEOUS↗
DOE OSTI2025

Laboratory Directed Research and Development Program: FY 2024 Completed Projects Report

Oak Ridge National Laboratory (ORNL) is the US Department of Energy’s (DOE’s) largest multiprogram science, technology, and energy laboratory. It possesses distinctive capabilities in a variety of fields, such as neutron science, computing, advanced materials, and nuclear science and technology. Using these capabilities, ORNL conducts basic and applied research and development (R&D) to support DOE’s overarching mission “to ensure America’s security and prosperity by addressing its energy, environmental and nuclear challenges through transformative science and technology solutions.” As a national resource, ORNL also applies its capabilities and skills to the specific needs of other federal agencies and customers through the DOE Strategic Partnership Projects (SPP) Program. Information about the laboratory and its programs is available on the ORNL website. The Laboratory Directed Research and Development (LDRD) Program at ORNL operates under the authority of DOE Order 413.2C, Laboratory Directed Research and Development,3 which establishes DOE’s requirements for the program while providing the laboratory director broad flexibility for program implementation. The LDRD Program funds are obtained through a charge to all laboratory programs. Although it represents a relatively small portion of the overall research budget, the LDRD Program plays an essential role in maintaining the laboratory’s ability to respond to national needs. The program allows ORNL to improve its distinctive capabilities and to enhance its ability to conduct cutting-edge R&D. In accordance with the DOE order, R&D projects funded through the LDRD Program at ORNL support the goals of • maintaining the scientific and technical vitality of the laboratory; • enhancing the laboratory’s ability to address future DOE missions; • fostering creativity and stimulating exploration of forefront areas of science and technology; • serving as a proving ground for new concepts in R&D; and • supporting high-risk, potentially high-value R&D. This report provides an overview of the LDRD Program at ORNL in FY 2024 and contains summaries of all the LDRD research projects that concluded between October 1, 2023, and September 30, 2024.

99 GENERAL AND MISCELLANEOUS↗
DOE OSTI2024

Site heterogeneity and broad surface-binding isotherms in modern catalysis: Building intuition beyond the Sabatier principle

Learning the science of heterogeneous catalysis and electrocatalysis always starts with the simple case of a flat, uniform surface with an ideal adsorbate. It has of course been recognized for a century that real catalysts are more complicated. For the increasingly complex catalysts of the 21st century, this Perspective argues that surface heterogeneity and non-ideal binding isotherms are central features, and their implications need to be incorporated in current thinking. A variety of systems are described herein where catalyst complexity leads to broad, non-Langmuirian surface isotherms for the binding of hydrogen atoms – and this occurs even for ideal, flat Pt(111) surfaces. Modern catalysis employs nanoscale materials whose surfaces have substantial step, edge, corner, impurity, and other defect sites, and they increasingly have both metallic and non-metallic elements M n X m , including metal oxides, chalcogenides, pnictides, carbides, doped carbons, etc. The surfaces of such catalysts are often not crystal facets of the bulk phase underneath, and they typically have a variety of potential active sites. Catalytic surfaces in operando are often non-stoichiometric, amorphous, dynamic, and impure, and often vary from one part of the surface to another. Understanding of the issues that arise at such nanoscale, multi-element catalysts is just beginning to emerge. Yet these catalysts are widely discussed using Brønsted/Bell-Evans-Polanyi (BEP) relations, volcano plots, Tafel slopes, the Butler-Volmer equation, and other linear free energy relations (LFERs), which all depend on the implicit assumption that the active sites are “similar” and that surface adsorption is close to ideal. These assumptions underly the ubiquitous intuition based on the Sabatier Principle, that the fastest catalysis will occur when key intermediates have free energies of adsorption that are not too strong nor too weak. Current catalysis research often aims to minimize the complexity of non-ideal isotherms through experimental and computational design (e.g., the use of single crystal surfaces), and these studies are the foundation of the field. In contrast, this Perspective argues that the heterogeneity of binding sites and binding energies is an inherent strength of these catalysts. Here, this diversity makes many nanoscale catalysts inherently a high-throughput screen wrapped in a tiny package. Only by making the heterogeneity part of the foundation of catalysis models, sorting the types of active sites and dissecting non-ideal binding isotherms, will modern catalysis learn to harness the inherent diversity of real catalysts. Controlling and exploiting diversity rather than avoiding it will help to optimize complex modern catalysts and catalytic conditions.

Mayer, James M.↗
DOE OSTI2020

China’s Bid To Lead Artificial Intelligence Chip Development within the Decade

In 2017, the Chinese government announced an ambitious, broad strategy to become the global leader in artificial intelligence (AI) theories, technologies, and applications by 2030 and indicated that China’s ability to indigenously produce cutting-edge AI chips would be integral to its success. China’s goal to become a dominant producer of AI hardware within a decade is a bold undertaking because China faces domestic chip production challenges and fierce competition from US chip producers. Our review of Chinese AI chip product lines shows various levels where Chinese companies compete with global firms in the production chain. China lacks a robust indigenous AI chip production infrastructure, and the United States and its allies are tightening controls on the supply of advanced semiconductor manufacturing equipment (SME) and electronic design automation (EDA) software to China. In contrast to their Chinese counterparts, US companies operate in all areas of AI chip production and have consistently driven the development of new chip designs. AI chip production is a costly endeavor, and China’s chip producers currently lack the customer base to lead global AI chip development by the 2030 goal. An expanded customer base may help China attract the necessary expertise to bolster its AI chip production chain. This would also help Chinese producers make their investments in the requisite technology economically sustainable. Chinese researchers continue to pursue promising next-generation AI chip designs, such as neuromorphic computing-based architectures, which in the next five to ten years could position Chinese companies to make or contribute to key innovations in the field.

74 ATOMIC AND MOLECULAR PHYSICS↗
DOE OSTI2021

Flattening the Radial Temperature Profile across the Transformational Challenge Reactor Core

The Transformational Challenge Reactor (TCR) program is demonstrating an agile development approach to advanced nuclear reactor design, which has traditionally utilized a linear design process. In leveraging artificial intelligence, additive manufacturing, advanced materials, and cutting-edge modeling and simulation, the TCR program aims to minimize the high cost and lengthy deployment timelines now standard in the nuclear industry. Within a relatively short period of time, a robust and mature advanced gas-cooled reactor was iteratively designed under the TCR program, using these cutting-edge technologies. The TCR is a 3 MWt gas-cooled microreactor fueled with uranium nitride (UN) tristructural isotropic (TRISO) fuel particles. Though manufactured via traditional means, these UN TRISO particles are loaded into additively manufactured silicon carbide (SiC) cans [4]. Once loaded with TRISO particles, the SiC cans are densified using a chemical vapor infiltration process. The additively manufactured SiC enables significantly more freedom in the design of the fuel form than could ever be achieved using traditionally manufactured SiC. The helium coolant, pressurized to 5 MPa, enters the core at 300°C and nominally exits it at 500°C. Typically, the most thermally limiting components in any reactor design are the fuel assemblies in the core center. To provide a wide thermal margin in these central fuel assemblies, the flow may be biased toward the center of the core to more effectively cool these fuel assemblies with more power deposition and flatten the core’s radial temperature distribution. An analytical fluid model of the TCR core was developed to explore methods for biasing the flow away from the cooler outer fuel assemblies and towards the hotter inner ones. Higher-fidelity models developed in STAR-CCM+ 2020.3.1, a computational fluid dynamics code, were then utilized to verify the analytical model’s findings.

22 GENERAL STUDIES OF NUCLEAR REACTORS↗
DOE OSTI2026

Cybersecurity at the Grid Edge: Protecting LA's Energy Systems in a Connected World

Join SoCalGas and the National Laboratory of the Rockies (NLR) for an insightful webinar on securing operational technology and energy systems. Los Angeles has the second-largest metro by population in the nation, making it vital to protect the energy infrastructure that powers day-to-day life. However, cybersecurity for energy systems is an immense challenge due to an increasing number of interconnected devices and stakeholders. While there was traditionally a limited need to secure energy infrastructure, the grid is only becoming smarter and more software-defined. Amid aging infrastructure and evolving cyber threats, the need to ensure the security of energy is at an all-time high. Together, SoCalGas and NLR are assessing the current state-of-the-art of the energy ecosystem to understand gaps in current best practices and technologies relevant to powering the city of Los Angeles.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2021

Phase-field modeling of the interactions between an edge dislocation and an array of obstacles

Obstacles, such as voids and precipitates, are prevalent in crystalline materials. They strengthen crystals by serving as barriers to dislocation glide. Here in this work, we develop a phase-field dislocation dynamics (PFDD) technique for investigating the interactions between dislocations and second-phase obstacles, which can be either voids or precipitates. The PFDD technique is constructed to account for elastic heterogeneity, elastic anisotropy, dissociation of the dislocation, and dislocation transmission across bicrystalline interfaces. Within the framework, we present a model for “pseudo-voids”, which are voids shearable by dislocations, in contrast to unphysical, unshearable voids in conventional phase-field dislocation formulations. We employ the PFDD technique to investigate the in-plane interactions between an edge dislocation and an array of nano-scale obstacles with different spacings. In this application, the interactions take place in glide planes of either a face-centered cubic (FCC) Cu or a body-centered cubic (BCC) Nb matrix, while the precipitates have a Cu 1-x Nb x composition, with x varying from 0.1 to 0.9. Our atomistic simulations find that the alloy precipitates can have an FCC, an amorphous, or a BCC phase, depending on the compositional ratio between Cu and Nb, i.e., value of x. Among all types of obstacles, the critical stresses for dislocation bypass are the highest for unshearable amorphous precipitates, followed by shearable crystalline precipitates, and then the pseudo-voids.

42 ENGINEERING↗
DOE OSTI2024

Phase separation, edge currents, and Hall effect for active matter with Magnus dynamics

Here we examine run-and-tumble disks in two-dimensional systems where the particles also have a Magnus component to their dynamics. For increased activity, we find that the system forms a motility-induced phase-separated (MIPS) state with chiral edge flow around the clusters, where the direction of the current is correlated with the sign of the Magnus term. The stability of the MIPS state is non-monotonic as a function of increasing Magnus term amplitude, with the MIPS region first extending down to lower activities followed by a break up of MIPS at large Magnus amplitudes into a gel-like state. We examine the dynamics in the presence of quenched disorder and a uniform drive and find that the bulk flow exhibits a drive-dependent Hall angle. This is a result of the side jump effect produced by scattering from the pinning sites and is similar to the behavior found for skyrmions in chiral magnets with quenched disorder.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2025

Representing Complex Systems as Graphs for Debugging and Predictive Maintenance-Preliminary Thoughts

Representing complex systems as graphs enables use of mathematical tools to identify faults or predict failures. Graph nodes correspond to individual modules or subsystems, and edges link coupled system parts. ‘Probes’ measure the node outputs, monitoring the system health for unexpected behavior. Assuming one cannot probe every point, within a system, the fault correlates to a region—not necessarily the specific location. Bayesian networks trained to understand fault patterns can accurately identify the source. The diagnostic tool described aides debugging by pinpointing system failure causes. For predictive maintenance, probe data develop probability distribution functions describing subsystem mean time to failure. Unit lifetime can be estimated through these probability distributions. Two approaches include using Bayesian classifiers to infer the system failure source and developing maintenance schedules by treating systems as collections of random variables. When failure behavior does not follow a closed form function, use of similarity models is proposed.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2023

Robust Real-Time Modeling of Distribution Systems with Data-Driven Grid-Wise Observability (Final Technical Report)

The overall objective of this project is to leverage existing and emerging sensor measurements to develop data-driven observability enhancement algorithms as well as robust state estimation and parameter identification techniques to enable real-time grid-wise monitoring and modeling of loads and distributed energy resources (DERs). The project resulted in a holistic framework with the following key components: 1) data-driven and machine learning-based grid-edge visibility enhancement, 2) robust branch-current-based state estimation (BCSE), and 3) robust real-time steady-state and dynamic-state modeling of loads and DERs. The research outcomes have provided utility companies better network visibility, higher-fidelity load/DER models, and more accurate assessment of DERs’ impacts, thus, facilitating the integration of renewable energy sources. The synergistic collaborative project among Iowa State University (ISU), Argonne National Laboratory (ANL), Electric Power Research Center (EPRC), SIEMENS Industry, Alliant Energy, Cedar Falls Utilities (CFU), and Maquoketa Valley Electric Cooperative (MVEC) to leverage the team’s extensive expertise and experience in power distribution systems, state estimation, online modeling and identification, and DER integrations. The proposed frameworks have been verified with industry adopted software and attempted integrated with existing tool wherever possible

24 POWER TRANSMISSION AND DISTRIBUTION↗
DOE OSTI2026

NeuroCoreX: An Open-Source FPGA-Based Spiking Neural Network Emulator with On-Chip Learning

Spiking Neural Networks (SNNs) are computational models inspired by the event-driven communication and connectivity patterns of biological neural circuits. They enable high energy efficiency and natural support for diverse architectures ranging from layered networks to small-world and graphstructured topologies. In this work, we introduce NeuroCoreX, an open-source, FPGA-based spiking neural network emulator that provides real-time, on-chip learning and flexible network organization. NeuroCoreX supports both feedforward sensory inputs streamed directly from sensors or PCs via UART and recurrent on-chip connectivity, enabling simultaneous processing and learning from external stimuli and internal network dynamics-capabilities rarely available in existing FPGA SNN platforms. The system implements a Leaky Integrate-and-Fire (LIF) neuron model with current-based synapses and supports pair-based STDP learning on both feedforward and recurrent synapses. A lightweight Python interface enables interactive configuration, live monitoring, weight read-back, and experiment control. Importantly, NeuroCoreX is tightly integrated with the SuperNeuroMAT simulator, allowing SNN models to be transferred seamlessly from software to hardware for hardware-in-the-loop development. By combining real-time plasticity, flexible connectivity, and an open-source VHDL implementation, NeuroCoreX provides an extensible and accessible platform for neuromorphic research, algorithm-hardware co-design, and energy-efficient edge intelligence.

Gautam, Ashish [ORNL]↗
DOE OSTI2025

Dry electrodes with a printed cellulose–graphene ink for low-profile strain sensors in electromyography

Dihydrolevoglucosenone, commonly known as Cyrene, is a renewable and fully biodegradable cellulose-waste derived, environmentally friendly solvent, presenting a non-toxic alternative to N-methyl-2-pyrrolidone (NMP). Currently, solution-based processing of graphene and other similar van der Waals solids favor toxic solvents such as NMP, limiting their use for biosensing. However, with the use of Cyrene, bio-compatible printable devices are possible, and studies have already demonstrated its use in temperature and other biosensing methods through screen-printing. Screen-printing unfortunately often requires masks that constrain the minimum acquirable feature size to be above hundreds of centimeters and wastes material, adding to process complexity and cost. Conversely, inkjet-printing is an attractive alternative for the maskless patterning of hierarchically assembled structures, with micron length scales attainable. Graphene's high conductivity positions it ideally for long-wear sensors such as dry electrodes or respiration monitors. Here, we demonstrate the potential of Cyrene-based graphene inks through few-layer inkjet printing on flexible substrates for the first time, to produce non-toxic conductors toward a strain-mediated mechanism for biosensing, used to detect bodily motion for wearable electronics. The challenges overcome in this study include engineering ink chemistry and printing parameters such that Cyrene's relatively high viscosity compared to typical inkjet solvents, still allows for droplet ejection in a conventional material printer, yielding well-resolved clean line-edges in contrast to other solvents that exhibit diffuse line-edges possibly from stray droplets and ink-splashing. Temperature-dependent transport measurements on the inkjet-printed Cyrene-based graphene films showed the conductivity to be largely temperature-invariant but at lower temperatures below 100 K, conductivity decreased, likely as a result of increased inter-membrane separation arising from thermal contraction. Additionally, temperature-dependent Raman spectroscopy showed the red-shift in the G-band, 2D-band and D-band peaks, as temperature increased. As a result, by validating flexion motion detection of the proximal interphalangeal joint demonstrated in this study, our work is the first of its kind to successfully additively manufacture inkjet-printed Cyrene-based graphene strain sensors on flexible substrates for bio-sensing and wearables.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2020

Hypernetwork science via high-order hypergraph walks

We propose high-order hypergraph walks as a framework to generalize graph-based network science techniques to hypergraphs. Edge incidence in hypergraphs is quantitative, yielding hypergraph walks with both length and width. Graph methods which then generalize to hypergraphs include connected component analyses, graph distance-based metrics such as closeness centrality, and motif-based measures such as clustering coefficients. We apply high-order analogs of these methods to real world hypernetworks, and show they reveal nuanced and interpretable structure that cannot be detected by graph-based methods. Lastly, we apply three generative models to the data and find that basic hypergraph properties, such as density and degree distributions, do not necessarily control these new structural measurements. Our work demonstrates how analyses of hypergraph-structured data are richer when utilizing tools tailored to capture hypergraph-native phenomena, and suggests one possible avenue towards that end.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2026

Using Explainable Artificial Intelligence to Predict Perovskite Solar Cell Electrical Metastability from Operando Photoluminescence Images in Accelerated Stress Testing

Metal halide perovskite (MHP) solar cells exhibit a metastable response to bias governed by coupled ionic–electronic processes, complicating the conventional reciprocity relation between luminescence intensity and device open-circuit voltage (V oc ). This limits the use of luminescence as a diagnostic for device screening or accelerated stress testing, motivating new approaches that can interpret photoluminescence (PL) signals under nonequilibrium conditions. From the artificial intelligence perspective, we develop an explainable deep learning framework that integrates convolutional neural networks (CNN), long short-term memory (LSTM) layers, and an attention mechanism to learn spatiotemporal features from operando photoluminescence PL image sequences. The model achieves a mean absolute error of ±0.027 V in predicting open-circuit voltage transients and reduces extreme-tail errors by up to 78% compared to physics-based reciprocity calculations. Gradient-weighted Class Activation Mapping (Grad-CAM) provides interpretability by highlighting physically meaningful regions such as electrode edges and emergent defect features. From the engineering application perspective, this framework enables accurate, contactless prediction of device V oc and identification of degradation-relevant features during accelerated aging of perovskite solar cells. This approach demonstrates how explainable AI can enhance operando diagnostics and reliability analysis in photovoltaic devices under nonequilibrium conditions.

14 SOLAR ENERGY↗
DOE OSTI2023

The INTERSECT Open Federated Architecture for the Laboratory of the Future

A federated instrument-to-edge-to-center architecture is needed to autonomously collect, transfer, store, process, curate, and archive scientific data and reduce human-in-the-loop needs with (a) common interfaces to leverage community and custom software, (b) pluggability to permit adaptable solutions, reuse, and digital twins, and (c) an open standard to enable adoption by science facilities world-wide. The Selfdriven Experiments for Science/Interconnected Science Ecosystem (INTERSECT) Open Architecture enables science breakthroughs using intelligent networked systems, instruments and facilities with autonomous experiments, “self-driving” laboratories, smart manufacturing and artificial intelligence (AI) driven design, discovery and evaluation. It creates an open federated architecture for the laboratory of the future using a novel approach, consisting of (1) science use case design patterns, (2) a system of systems architecture, and (3) a microservice architecture.

Engelmann, Christian↗
DOE OSTI2025

Using a Large Language Model as a Building Block to Generate Usable Validation and Verification Suite for OpenMP

In the HPC area, both hardware and software move quickly. Often new hardware is developed and deployed, the corresponding software stack, including compilers and other tools, are under active development while leading edge software developers are working to port and tune their applications, all at the same time. While the software ecosystem is in flux, one of the key challenges for users is obtaining insight into the state of implementation of key features in the programming languages and models their applications are using – whether they have been implemented, and whether the implementation conforms to the specification, especially for newly implemented features (less tested by widespread use). OpenMP is one of the most prominent shared memory programming models used for on-node programming in HPC. With the shift towards accelerators (such as GPUs and FPGAs) and heterogeneous programming OpenMP features are getting more complex. It is natural to ask whether generative AI approaches, and large language models (LLMs) in particular, can help in producing validation and verification test suites to allow users better and faster insights into the availability and correctness of OpenMP features of interest. In this work, we explore the use of ChatGPT-4 to generate a suite of tests for OpenMP features. We have chosen a set of directives and clauses, a total of 78 combinations, which first appeared in OpenMP 3.0 (released in May 2008) but are also relevant for accelerators. We prompted ChatGPT to generate tests in the C and Fortran languages, for both host (CPU) and device (accelerator). On the Summit super-computer using the GNU implementation, we found that, of the 78 generated tests 67 C tests and 43 Fortran tests compiled successfully and fewer than those executed to completion. On further analysis we show that not all generated tests are valid. We document the process, results, and provide detailed analysis regarding the quality of tests generated. With the aim of providing input to a production quality validation and verification suite, we manually implement the corrections required to make the tests valid according to the current OpenMP specification. We quantify this effort as small, medium, or large, and record the lines of code changed to correct the invalid tests. With the corrected tests we validate recent implementations from HPE, AMD, and GNU on the Frontier supercomputer. Our experiment and subsequent analysis show that although LLMs are capable of producing HPC specific codes, they are limited by their understanding of the deeper semantics and restrictions of programming models such as OpenMP. Unsurprisingly more commonly used features have better support, while some OpenMP 3.0 directives such as sections and tasking are not universally supported on accelerators. We demonstrate that successful compilation and execution to completion are inadequate metrics for evaluating generated code and that, at this time, commodity LLMs require expert intervention for code verification. This points to gaps in the training data that is currently available for HPC. We demonstrate that with "small" effort 37% of generated invalid C tests and 63% of generated invalid Fortran tests could be corrected. This improves productivity of test generation as we circumvent writing from scratch and the common programming errors associated with it.

Pophale, Swaroop [ORNL] (ORCID:0000000185446367)↗