Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Advanced Supercomputing”

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 253 records · Page 14

Near Real-Time Two and Three-Dimensional Electrical Geophysical Monitoring of Natural and Engineered Processes Associated with Subsurface Remediation - 16052

Legacy soil and groundwater contamination associated with nuclear weapons production is projected to be one of the largest cleanup liabilities to the U.S. Department of Energy for the foreseeable future. Remediation of subsurface contaminants is complicated by the combined effects of geochemical heterogeneity, geological heterogeneity, and the difficulty and expense of adequate borehole access. Each of these complications contributes to uncertainty in the distribution of contaminants and the performance of in situ treatments, which increases risk to human health and the environment and cost to closure in general. Electrical resistivity tomography (ERT) is a method of remotely imaging the electrical properties of the subsurface, which are governed by both geochemical and geological structure, thereby providing useful proxies for understanding contaminant distribution and the behavior in situ remediation processes. Recent advancements in both data collection and data processing capabilities are enabling ERT monitoring to be executed and provided to site operators in near real-time. To summarize the process, ERT surveys are rapidly and continuously collected during a time-sensitive operation such as an amendment injection. At the completion of each survey, data are transferred by wireless internet to offsite supercomputing resources for parallel tomographic inversion. Inversion results are then transferred back to onsite operators and/or other locations for visualization. In this paper, we demonstrate the utility of time-lapse ERT for monitoring fluid transport (both gas and liquid phase). We then demonstrate two real-time imaging applications for monitoring both natural and engineered processes at the Hanford Site within the vadose and the saturated zones. We expect this new capability for autonomous real-time imaging to enable ERT to be used as a cost-effective, short-term rapid feedback mechanism for guiding subsurface remediation injections, in addition to longer term monitoring of post-injection performance and environmental impact, all using the same electrode array and instrumentation. (authors)

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗

Free Energy Landscapes of Membrane Transport Proteins

Molecular dynamics (MD) is now a widespread tool for investigating biochemical and biomolecular systems, its prevalence, in no small part, being due to significant advances in computational hardware over the last few decades. Even non-experts can now routinely use MD for protein and nucleic acid structure refinement, studying conformational switching events, and investigating solvation effects due to ligand/drug binding. Nonetheless, the vast majority of simulations being done today utilize only rudimentary algorithmic approaches – so-called “brute force” MD – which only permit access to a small fraction of what the approach has to offer. This is largely because the core algorithm is simple, while advanced techniques can require everything from more sophisticated models and data structures to complicated on-the-fly analysis. A core goal of this early science project is to bring one such advanced MD approach into the broader arena of high-performance computing.In particular, the powerful method of constant-pH MD has long been a relegated to experienced users and specialized model systems. In this work we develop and implement a constant-pH MD algorithm in the NAMD simulation engine making it suitable for deployment on large-scale, next-generation supercomputers as well as ambitious, cutting-edge biological applications. We report,for the first time, constant-pH simulations of a membrane transport protein and use the results to analyze its free energy landscape for ion-selectivity.

59 BASIC BIOLOGICAL SCIENCES↗

Towards an understanding of complex chemical systems: A scalable molecular simulation approach

Energy production and security are among the major challenges facing the nation. Technological breakthroughs must occur for production, storage, and use of energy with performance levels far beyond what is now possible. Only with these breakthroughs will we be able to ensure energy independence, environmental sustainability, and continued economic opportunities. To accomplish our goals will require us to have the ability to design and control matter at the molecular level of electrons and nuclei. Through quantum mechanics we can understand the chemical, electrical, and physical properties of materials. Computational chemistry allows us to examine questions that are not possible or are very difficult to examine in the laboratory. Only with the ability to carry out the computations required for the explicit treatment of quantum mechanics can we hope to pave the way for highly efficient energy technologies needed for the future. Increasingly powerful computational capabilities allows for increasing sophisticated analyses. Scientists at the Pacific Northwest National Laboratory and the University of Zurich are conducting research and developing state-of-the-art molecular simulation tools, combined with increasingly powerful supercomputers, to enable a greater understanding and manipulation of complex chemical processes. Researchers are applying efficient algorithms to obtain the dynamics of atoms and nuclei interacting under the laws of quantum mechanics to develop a new understanding of chemical reactions in solutions and at interfaces, especially as they relate to hydrogen storage and catalysis. The insights gained from these calculations will enable technological advances capable of addressing the U.S. Department of Energy’s (DOE) energy mission and keeping the US at the forefront of technology development.

Mundy, Christopher J.↗

Fluid learning: Mimicking brain computing with neuromorphic nanofluidic devices

Relentlessly rising energy demands in computing call for rethinking hardware paradigms with energy efficiency in mind. Nature’s example—the brain—raises the question: How can these natural computers achieve remarkable feats with minimal energy compared to supercomputers? Neuromorphic computing mimics the brain’s principles, but current neuromorphic concepts using electronic components face scalability and their own power consumption challenges. A potentially revolutionary approach is emerging: computing with ion transport in water through nanochannels. This field offers energy-efficient possibilities by imitating brain-like information processing with different types of ions as carriers. Finally, the goal is to converge advanced nanoscale architectures with brain-inspired efficiency, heralding a new era of computing.

77 NANOSCIENCE AND NANOTECHNOLOGY↗

Scientific Computing Strategic Plan for the Idaho National Laboratory

Scientific computing is a critical foundation of modern science. Without innovations in the field of computational science, the essential missions of the Department of Energy (DOE) would go unrealized. Taking a leadership role in such innovations is Idaho National Laboratory’s (INL’s) challenge and charge, and is central to INL’s ongoing success. Computing is an essential part of INL’s future. DOE science and technology missions rely firmly on computing capabilities in various forms. Modeling and simulation, fueled by innovations in computational science and validated through experiment, are a critical foundation of science and engineering. Big data analytics from an increasing number of widely varied sources is opening new windows of insight and discovery. Computing is a critical tool in education, science, engineering, and experiments. Advanced computing capabilities in the form of people, tools, computers, and facilities, will position INL competitively to deliver results and solutions on important national science and engineering challenges. A computing strategy must include much more than simply computers. The foundational enabling component of computing at many DOE national laboratories is the combination of a showcase like data center facility coupled with a very capable supercomputer. In addition, network connectivity, disk storage systems, and visualization hardware are critical and generally tightly coupled to the computer system and co located in the same facility. The existence of these resources in a single data center facility opens the doors to many opportunities that would not otherwise be possible.

97 MATHEMATICS AND COMPUTING↗

Exascale Workshop Panel Report Meeting

The Exascale Review Panel consists of 12 scientists and engineers with experience in various aspects of high-performance computing and its application, development, and management. The Panel hear presentations by several representatives of the workshops and town meetings convened over the past few years to examine the need for exascale computation capability and the justification for a U.S. Department of Energy (DOE) program to develop such capability. This report summarizes information provided by the presenters and substantial written reports to the Panel in advance of the meeting in Washington D.C. on January 19-20, 2010. The report also summarizes the Panel's conclusions with regard to the justification of a DOE-led exascale initiative.

97 MATHEMATICS AND COMPUTING↗

Argonne National Laboratory Annual Report of Laboratory Directed Research and Development Program Activities FY 2012

Our plans to upgrade the Advanced Photon Source are proceeding on schedule, and we hope to receive CD-2 approval sometime in spring 2013. When complete, the APS Upgrade will equip this vital facility for continued scientific leadership in the 21st century, providing the 5,000-plus members of our user community with greatly increased time, energy and spatial resolution. We also have successfully installed “Mira,” the new IBM Blue Gene/Q system at the Argonne Leadership Computing Facility. The world’s fourth-fastest supercomputer, Mira is now providing leading-edge computational power across all scientific disciplines. I am also pleased to report that Argonne’s new Energy Sciences Building is proceeding on schedule; we expect it to be finished and ready for occupancy by June 30, 2013. When complete, the ESB will be one of the most modern and most flexible energy labs in the world, creating a new home for researchers focused on a wide range of energy science challenges. Research is under way at the new multidisciplinary Institute for Molecular Engineering (IME), a joint program of the University of Chicago and Argonne. Over the past year, four new faculty members, including IME’s director, – all pre-eminent researchers who hold joint appointments at Argonne – have joined the IME. We look forward to working in collaboration with the IME to find new ways to address the interfaces between molecular-level science and powerful new technologies. In FY 2012, we continued support for the final year of the “Beyond Lithium-Ion” Director’s Grand Challenge LDRD project. That project enabled significant work on the performance of lithium-based energy storage systems, pointing the way to new high-energy-density lithium-air batteries. I believe this Grand Challenge contributed significantly to our successful bid to win designation of the Argonne-based Joint Center for Energy Storage Research (JCESR) as the DOE’s new Batteries and Energy Storage Hub. JCESR, which brings together a world-class “dream team” of national laboratories, universities and private sector partners, strongly benefited from LDRD program investments, both in FY2012 and in previous years. As we move forward in FY2013, the Director’s Grand Challenge is now focused on new ways to address the challenges of “big data” – the acquisition, storage, transfer, and analysis of both experimental and simulated data at scales approaching the limits of our current computational capacities. In managing Argonne’s LDRD program, we have adhered to the requirements of DOE Order 413.2b and associated guidelines. Our program management continually strives for increased efficiency. In addition to meeting all reporting requirements during FY 2012, our LDRD Program Office continues to implement intranet-based process enhancements to better serve all LDRD stakeholders, including our DOE Site Office colleagues. As the following individual project reports demonstrate, Argonne’s researchers are pursuing projects at the leading edge of science and engineering, and making significant contributions to the advancement of our laboratory’s major strategic initiatives. Evidence of the continuing strength of this program is found in the follow-on sponsorship won by many of these research projects, as well as in the numbers of publications, invited presentations, inventions, and new staff hires that are direct results of LDRD. LDRD enables exploration of new scientific directions and makes it possible to adapt quickly to a changing landscape with unexpected opportunities, while enhancing the capabilities of our researchers and our scientific facilities. For these reasons, LDRD remains an invaluable resource to address the future science and technology needs of DOE, and of our nation.

96 KNOWLEDGE MANAGEMENT AND PRESERVATION↗

Secure Federated Learning Across Heterogeneous Cloud and High-Performance Computing Resources: A Case Study on Federated Fine-Tuning of LLaMA 2

Federated learning enables multiple data owners to collaboratively train robust machine learning models without transferring large or sensitive local datasets by only sharing the parameters of the locally trained models. Here, in this article, we elaborate on the design of our Advanced Privacy-Preserving Federated Learning (APPFL) framework, which streamlines end-to-end secure and reliable federated learning experiments across cloud computing facilities and high-performance computing resources by leveraging Globus Compute, a distributed function as a service platform, and Amazon Web Services. We further demonstrate the use case of APPFL in fine-tuning an LLaMA 2 7B model using several cloud resources and supercomputers.

97 MATHEMATICS AND COMPUTING↗

MapReduce SVM Game

Despite technological advances making computing devices faster, smaller, and more prevalent in today's age, data generation and collection has outpaced data processing capabilities. Simply having more compute platforms does not provide a means of addressing challenging problems in the big data era. Rather, alternative processing approaches are needed and the application of machine learning to big data is hugely important. The MapReduce programming paradigm is an alternative to conventional supercomputing approaches, and requires less stringent data passing constrained problem decompositions. Rather, MapReduce relies upon defining a means of partitioning the desired problem so that subsets may be computed independently and recom- bined to yield the net desired result. However, not all machine learning algorithms are amenable to such an approach. Game-theoretic algorithms are often innately distributed, consisting of local interactions between players without requiring a central authority and are iterative by nature rather than requiring extensive retraining. Effectively, a game-theoretic approach to machine learning is well suited for the MapReduce paradigm and provides a novel, alternative new perspective to addressing the big data problem. In this paper we present a variant of our Support Vector Machine (SVM) Game classifier which may be used in a distributed manner, and show an illustrative example of applying this algorithm.

97 MATHEMATICS AND COMPUTING↗

Identification of stable adsorption sites and diffusion paths on nanocluster surfaces: an automated scanning algorithm

The diverse coordination environments on the surfaces of discrete, three-dimensional (3D) nanoclusters contribute significantly to their unique catalytic properties. Identifying the numerous adsorption sites and diffusion paths on these clusters is however tedious and time-consuming, especially for large, asymmetric nanoclusters. Here, we present a simple, automated method for constructing approximate 2D potential energy surfaces for the adsorption of atomic species on the surfaces of 3D nanoclusters with minimal human intervention. These potential energy surfaces fully characterize the important adsorption sites and diffusion paths on the nanocluster surfaces with accuracies similar to current approaches and at comparable computational cost. Our method can treat complex nanoclusters, such as alloy nanoclusters, and accounts for cluster relaxation and adsorbate-induced reconstruction, important for obtaining accurate energetics. Moreover, its highly parallelizable nature is ideal for modern supercomputer architectures. We showcase our method using two clusters: Au 18 and Pt 55 . For Au 18 , diffusion of atomic hydrogen between the most stable sites occurs via non-intuitive paths, underlining the necessity of exploring the complete potential energy surface. By enabling the rapid and unbiased assessment of adsorption and diffusion on large, complex nanoclusters, which are particularly difficult to handle manually, our method will help advance materials discovery and the rational design of catalysts.

36 MATERIALS SCIENCE↗

WorkflowHub: a registry for computational workflows

The rising popularity of computational workflows is driven by the need for repetitive and scalable data processing, sharing of processing know-how, and transparent methods. As both combined records of analysis and descriptions of processing steps, workflows should be reproducible, reusable, adaptable, and available. Workflow sharing presents opportunities to reduce unnecessary reinvention, promote reuse, increase access to best practice analyses for non-experts, and increase productivity. In reality, workflows are scattered and difficult to find, in part due to the diversity of available workflow engines and ecosystems, and because workflow sharing is not yet part of research practice. WorkflowHub provides a unified registry for all computational workflows that links to community repositories, and supports both the workflow lifecycle and making workflows findable, accessible, interoperable, and reusable (FAIR). By interoperating with diverse platforms, services, and external registries, WorkflowHub adds value by supporting workflow sharing, explicitly assigning credit, enhancing FAIRness, and promoting workflows as scholarly artefacts. The registry has a global reach, with hundreds of research organisations involved, and more than 800 workflows registered.

97 MATHEMATICS AND COMPUTING↗

Demonstrating Cross-Facility Data Processing at Scale with Laue Microdiffraction

In February and April 2023 live, at-scale data processing demonstrations were conducted between the Advanced Photon Source (APS), a synchrotron light source, and the Argonne Leadership Computing Facility (ALCF). These tests were run as part of a novel beamline technique: coded aperture laue micro-diffraction. This technique requires a significant amount of compute to decode appeture patterns embedded in the detector stream. An autonomous system was able to send data to ALCF during an experiment, utilize 50 nodes of the Polaris supercomputer to process 6-12 hour scans, and return the data back to the APS within 12-15 minutes behind the detector. With scan points arriving every 72 seconds, the system kept up with the beamline, potentially enabling in-experiment analysis. The data processing system utilizes Globus infrastructure and an on-demand queue to dynamically acquire nodes on Polaris. The underlying reconstruction algorithms were parallelized via MPI and accelerated with custom CUDA kernels.

Prince, Michael↗

Performance of asynchronous optimized Schwarz with one-sided communication

In asynchronous iterative methods on distributed-memory computers, processes update their local solutions using data from other processes without an implicit or explicit global synchronization that corresponds to advancing the global iteration counter. Here in this work, we test the asynchronous optimized Schwarz domain-decomposition iterative method using various one-sided (remote direct memory access) communication schemes with passive target completion. The results show that when one-sided communication is well-supported, the asynchronous version of optimized Schwarz can outperform the synchronous version even for perfectly balanced partitionings of the problem on a supercomputer with uniform nodes.

97 MATHEMATICS AND COMPUTING↗

State of the art in electromagnetic modeling for the Compact Linear Collider

SLAC's Advanced Computations Department (ACD) has developed the parallel 3D electromagnetic time-domain code T3P for simulations of wakefields and transients in complex accelerator structures. T3P is based on state-of-the-art Finite Element methods on unstructured grids and features unconditional stability, quadratic surface approximation and up to 6th-order vector basis functions for unprecedented simulation accuracy. Optimized for large-scale parallel processing on leadership supercomputing facilities, T3P allows simulations of realistic 3D structures with fast turn-around times, aiding the design of the next generation of accelerator facilities. Applications include simulations of the proposed two-beam accelerator structures for the Compact Linear Collider (CLIC) - wakefield damping in the Power Extraction and Transfer Structure (PETS) and power transfer to the main beam accelerating structures are investigated.

43 PARTICLE ACCELERATORS↗

The Scientific Impact of the Exascale Computing Project

The recent arrival of the Frontier Supercomputer at Oak Ridge National Laboratory officially marked the dawn of the exascale computing era. Its successful deployment coincided with the culmination of the U.S. Department of Energy Exascale Computing Project (ECP), an ambitious, complex, and risky research and development effort that integrated contributions from a broad and diverse subset of the high-performance computing community. The success of ECP will ultimately be judged by the scientific and engineering advances that it enabled. In conclusion, this Special Issue is focused on showcasing early successes in the use of exascale resources to enable breakthroughs in key areas of science in engineering.

97 MATHEMATICS AND COMPUTING↗

Deploying Darter A Cray XC30 System

TheUniversityofTennessee,KnoxvilleacquiredaCrayXC30 supercomputer, called Darter, with a peak performance of 248.9 Ter- aflops. Darter was deployed in late March of 2013 with a very aggressive production timeline - the system was deployed, accepted, and placed into production in only 2 weeks. The Spring Experiment for the Center for Analysis and Prediction of Storms (CAPS) largely drove the accelerated timeline, as the experiment was scheduled to start in mid-April. The Consortium for Advanced Simulation of Light Water Reactors (CASL) project also needed access and was able to meet their tight deadlines on the newly acquired XC30. Darter s accelerated deployment and op- erations schedule resulted in substantial scientific impacts within the re- search community as well as immediate real-world impacts such as early severe tornado warnings

Fahey, Mark R↗

Multi-task Parallelism for Robust Pre-training of Graph Foundation Models on Multi-source, Multi-fidelity Atomistic Modeling Data

Graph foundation models using graph neural networks promise sustainable, efficient atomistic modeling. To tackle challenges of processing multi-source, multi-fidelity data during pre-training, recent studies employ multi-task learning, in which shared message passing layers initially process input atomistic structures regardless of source, then route them to multiple decoding heads that predict data-specific outputs. This approach stabilizes pre-training and enhances a model’s transferability to unexplored chemical regions. Preliminary results on approximately four million structures are encouraging, yet questions remain about generalizability to larger, more diverse datasets and scalability on supercomputers. We propose a multi-task parallelism method that distributes each head across computing resources with GPU acceleration. Implemented in the open-source HydraGNN architecture, our method was trained on over 24 million structures from five datasets and tested on the Perlmutter, Aurora, and Frontier supercomputers, demonstrating efficient scaling on all three highly heterogeneous super-computing architectures.

Lupo Pasini, Massimiliano [ORNL] (ORCID:0000000249↗

Simulating Billion-Task Parallel Programs

In simulating large parallel systems, bottom-up approaches exercise detailed hardware models with effects from simplified software models or traces, whereas top-down approaches evaluate the timing and functionality of detailed software models over coarse hardware models. Here, we focus on the top-down approach and significantly advance the scale of the simulated parallel programs. Via the direct execution technique combined with parallel discrete event simulation, we stretch the limits of the top-down approach by simulating message passing interface (MPI) programs with millions of tasks. Using a timing-validated benchmark application, a proof-of-concept scaling level is achieved to over 0.22 billion virtual MPI processes on 216,000 cores of a Cray XT5 supercomputer, representing one of the largest direct execution simulations to date, combined with a multiplexing ratio of 1024 simulated tasks per real task.

Perumalla, Kalyan S↗