Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Computational science”

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

Lincoln Advanced Science and Engineering Reinforcement (LASER) program

Lincoln University, under the Lincoln Advanced Science and Engineering Reinforcement (LASER) Program, has identified and successfully recruited over 100 students for majors in technical fields. To date, over 70 percent of these students have completed or will complete technical degrees in engineering, physics, chemistry, and computer science. Of those completing the undergraduate degree, over 40 percent have gone on to graduate and professional schools. This success is attributable to well planned approaches to student recruitment, training, personal motivation, retention, and program staff. Very closely coupled to the above factors is a focus designed to achieve excellence in program services and student performance. Future contributions by the LASER Program to the pool of technical minority graduates will have a significant impact. This is already evident from the success of the students that began the first year of the program. With program plans to refine many of the already successful techniques, follow-on activities are expected to make even greater contributions to the availability of technically trained minorities. For example, undergraduate research exposure, broadened summer, and co-op work experiences will be enhanced.

Williams, Willie E.↗

Neural Network Machine Learning and Dimension Reduction for Data Visualization

Neural network machine learning in computer science is a continuously developing field of study. Although neural network models have been developed which can accurately predict a numeric value or nominal classification, a general purpose method for constructing neural network architecture has yet to be developed. Computer scientists are often forced to rely on a trial-and-error process of developing and improving accurate neural network models. In many cases, models are constructed from a large number of input parameters. Understanding which input parameters have the greatest impact on the prediction of the model is often difficult to surmise, especially when the number of input variables is very high. This challenge is often labeled the "curse of dimensionality" in scientific fields. However, techniques exist for reducing the dimensionality of problems to just two dimensions. Once a problem's dimensions have been mapped to two dimensions, it can be easily plotted and understood by humans. The ability to visualize a multi-dimensional dataset can provide a means of identifying which input variables have the highest effect on determining a nominal or numeric output. Identifying these variables can provide a better means of training neural network models; models can be more easily and quickly trained using only input variables which appear to affect the outcome variable. The purpose of this project is to explore varying means of training neural networks and to utilize dimensional reduction for visualizing and understanding complex datasets.

Liles, Charles A.↗

Detecting Outliers in Network Transfers with Feature Extraction

Reliable file transfers are essential for successful science computations and experiments that need large complex data files to be moved across long distances. Networks use protocols such as TCP and UDP to support file transfers, and their performance degrades under packet losses and duplications, thereby adversely affecting science applications. Hence, scientists and engineers often monitor network statistics to find and repair network problems that cause such degradations.Outliers or anomalies can be detected using statistical and machine learning methods that highlight abnormal behaviors. However, without knowing what is normal, it is very difficult and subjective to determine the anomalies. In this paper, we investigate statistical and machine learning approaches to extract unknown or unsupervised features from TCP and perfSONAR data sets, with the aim to identify anomalies such as packet loss, duplication and retransmission sequencing that affect file transfer performance. Simple statistical and feature extraction methods (e.g. PCA) have shown that detecting anomalies is simpler if labeled data sets of normal behavior are available. While perfSONAR logs record losses,they do not detect duplications and reordering. We show that PCA applied to TCP statistics is able to determine abnormal behavior in the above cases. In particular, our results show that simple feature extraction techniques provide building blocks for more advanced network monitoring tools that can quickly flag anomalies using simple measurements, thereby helping to build confidence in file transfer reliability.

Rao, Nageswara S.↗

Visual Analytics: Building a Vibrant and Resilient National Science

Five years after the science of visual analytics was formally established, we attempt to use two different studies to assess the current state of the community and evaluate the progress the community has made in the past few years. The first study involves a comparison analysis of intellectual and scholastic accomplishments recently made by the visual analytics community. The second one aims to measure the degree of community reach and internet penetration of visual-analytics-related resources. This paper describes our efforts to harvest the study data, conduct analysis, and make interpretations based on parallel comparisons with five other established computer science areas.

97 MATHEMATICS AND COMPUTING↗

The role of machine learning in scientific workflows

Machine learning (ML) is being applied in a number of everyday contexts from image recognition, to natural language processing, to autonomous vehicles, to product recommendation. In the science realm, ML is being used for medical diagnosis, new materials development, smart agriculture, DNA classification, and many others. In this article, we describe the opportunities of using ML in the area of scientific workflow management. Scientific workflows are key to today’s computational science, enabling the definition and execution of complex applications in heterogeneous and often distributed environments. We describe the challenges of composing and executing scientific workflows and identify opportunities for applying ML techniques to meet these challenges by enhancing the current workflow management system capabilities. We foresee that as the ML field progresses, the automation provided by workflow management systems will greatly increase and result in significant improvements in scientific productivity.

97 MATHEMATICS AND COMPUTING↗

Challenges for monitoring and data analytics in a leadership public data repository

The availability and disposition of data has assumed increasing importance in large-scale computational science. Data repositories are evolving to meet new classes of requirements: compliance with government access guidelines, support for reproducibility of experimental results, and long-term availability of data products. The Constellation public data repository at the Oak Ridge Leadership Computing Facility faces these issues while being situated in one of the most productive data centers in the world. While monitoring and operational data analysis are ingrained in the operation of the OLCF’s large-scale high performance computing platforms, data repositories do not have this history of support. Problems faced by Constellation range from data size (over 7 petabytes in current holdings) to analytic complexity (detailed curation is both absolutely necessary for many data sets and absolutely impossible for humans to accomplish in any practical manner) to deployment environment (OLCF storage resources are oriented toward the needs of the compute platforms). In this paper we describe some of the challenges for collecting monitoring and analytic data from a leadership public data repository. We also discuss various strategies we are pursuing in order to address these challenges, from manual data collection to plans for introducing machine learning-based curatorial techniques.

Widener, Patrick [ORNL] (ORCID:0000000258820816)↗

Assessing the role of mini-applications in predicting key performance characteristics of scientific and engineering applications

Computational science and engineering application programs are typically large, complex, and dynamic, and are often constrained by distribution limitations. As a means of making tractable rapid explorations of scientific and engineering application programs in the context of new, emerging, and future computing architectures, a suite of miniapps has been created to serve as proxies for full scale applications. Each miniapp is designed to represent a key performance characteristic that does or is expected to significantly impact the runtime performance of an application program. In this paper we introduce a methodology for assessing the ability of these miniapps to effectively represent these performance issues. We applied this methodology to four miniapps, examining the linkage between them and an application they are intended to represent. Herein we evaluate the fidelity of that linkage. This work represents the initial steps required to begin to answer the question, ''Under what conditions does a miniapp represent a key performance characteristic in a full app?''

97 MATHEMATICS AND COMPUTING↗

What to Do Until the Money Runs Out: A Refinement Framework for Cognitive Engineering in the Real World

A case study is presented to illustrate some of the problems of applying cognitive science to complex human-machine systems. Disregard for facts about human cognition often undermines the safety, reliability, and cost-effectiveness of complex systems. Yet single-point methods (for example, better user-interface design), whether rooted in computer science or in experimental psychology, fall far short of addressing systems-level problems in a timely way using realistic resources. A model-based methodology is proposed for organizing and prioritizing the cognitive engineering effort, focusing appropriate expertise on major problems first, then moving to more sophisticated refinements if time and resources permit. This case study is based on a collaborative effort between the Human Factors Division at NASA-Ames and the Spaceborne Imaging Radar SIR-C/X-Band Synthetic Aperture Radar (SIR-C/X-SAR) Project at the Jet Propulsion Laboratory (JPL), California institute of Technology. The first SIR-C/X-SAR Shuttle mission flew successfully in April, 1994. A series of such missions is planned to provide radar data to study Earth's ecosystems, climatic and geological processes, hydrologic cycle, and ocean circulation. In addition to JPL and NASA personnel, the SIR-C/X-SAR operations team included Scientists and engineers from the German and Italian space agencies.

Shafto, Michael G.↗

A Multiagent Modeling Environment for Simulating Work Practice in Organizations

In this paper we position Brahms as a tool for simulating organizational processes. Brahms is a modeling and simulation environment for analyzing human work practice, and for using such models to develop intelligent software agents to support the work practice in organizations. Brahms is the result of more than ten years of research at the Institute for Research on Learning (IRL), NYNEX Science & Technology (the former R&D institute of the Baby Bell telephone company in New York, now Verizon), and for the last six years at NASA Ames Research Center, in the Work Systems Design and Evaluation group, part of the Computational Sciences Division (Code IC). Brahms has been used on more than ten modeling and simulation research projects, and recently has been used as a distributed multiagent development environment for developing work practice support tools for human in-situ science exploration on planetary surfaces, in particular a human mission to Mars. Brahms was originally conceived of as a business process modeling and simulation tool that incorporates the social systems of work, by illuminating how formal process flow descriptions relate to people s actual located activities in the workplace. Our research started in the early nineties as a reaction to experiences with work process modeling and simulation . Although an effective tool for convincing management of the potential cost-savings of the newly designed work processes, the modeling and simulation environment was only able to describe work as a normative workflow. However, the social systems, uncovered in work practices studied by the design team played a significant role in how work actually got done-actual lived work. Multi- tasking, informal assistance and circumstantial work interactions could not easily be represented in a tool with a strict workflow modeling paradigm. In response, we began to develop a tool that would have the benefits of work process modeling and simulation, but be distinctively able to represent the relations of people, locations, systems, artifacts, communication and information content.

Sierhuis, Maarten↗

Optimization of a Parallel Ocean General Circulation Model

Global climate modeling is one of the grand chalenges of computational science, and ocean modeling plays an important role in both understanding the current climatic conditions and predicting the future climate change.

Optimization Ocean Modeling Parallel Computations↗

Debugging Computer Code [Slides]

Debugging? Klingons do not debug. Our software does not coddle the weak. For the rest of us, finding and removing defects in code is a fundamental programming skill. Besides preventing code from performing its intended function, bugs destroy users’ confidence in the reliability of code even when no errors are obvious. A through discussion of debugging techniques could easily fill a semester course in computer science. We have an hour here today. I will take that time to discuss broad principles rather than fine points of debugging.

97 MATHEMATICS AND COMPUTING↗

The Atmospheric Channels of GLAS: Near Real-Time Global Lidar Remote Sensing of Clouds and Aerosols from Space

The Geoscience Laser Altimeter System (GLAS) will be placed into orbit in 2001 aboard the Ice, Cloud and Land Elevation Satellite (ICESat). From its nearly polar orbit (94 degree inclination), GLAS will provide continuous global measurements of the vertical distribution of clouds and aerosols while simultaneously providing high accuracy topographic profiling of surface features. During the mission, which is slated to last 3 to 5 years, the data collected by GLAS will be in near-real time to produce level 1 and 2 data products at the NASA GLAS Science Computing Facility (SCF) at Goddard Space Flight Center in Greenbelt, Maryland. The atmospheric products include cloud and aerosol layer heights, planetary boundary layer depth, polar stratospheric clouds and thin cloud and aerosol optical depth. These products will be made available to the science community within days of their creation. The processing algorithms must be robust, adaptive, efficient, and clever enough to run autonomously for the widely varying atmospheric conditions that will be encountered. This paper presents an overview of the GLAS atmospheric data products and briefly discusses the design of the processing algorithms.

Palm, Stephen P.↗

Advanced Simulation and Computing: ASC FY24 Implementation Plan

The DOE National Nuclear Security Administration (NNSA) Stockpile Stewardship Program (SSP) is an integrated technical program for maintaining the safety, security, and reliability of the U.S. nuclear stockpile. The SSP incorporates nuclear test data, computational modeling and simulation, and experimental facilities to advance understanding of nuclear weapons. The suite of data analyzed comes from activities including previous nuclear tests, stockpile surveillance, experimental research, and development and engineering programs. This integrated national program requires the continued use of experimental facilities and the computational capabilities to support the SSP missions. These component parts, in addition to an appropriately scaled production capability, enable NNSA to support stockpile requirements. The ultimate goal of the SSP, and thus of the Advanced Simulation and Computing (ASC) program, is to ensure that the U.S. maintains a safe, secure, and effective strategic deterrent. The ASC program is a cornerstone of the SSP, providing simulation capabilities and computational resources to support the annual stockpile assessment and certification process, study advanced nuclear weapons design and manufacturing processes, analyze accident scenarios and weapons aging, and provide the tools to enable stockpile Life Extension Programs (LEPs) and the resolution of Significant Finding Investigations (SFIs). This work requires a balance of resources, including technical staff, hardware, simulation software, and computer science solutions. The ASC program focuses on increasing the predictive capabilities in a three-dimensional (3D) simulation environment while maintaining support to the SSP. The Program continues to improve its unique tools for understanding and solving progressively more difficult stockpile problems (sufficient resolution, dimensionality, and scientific details), and quantifying critical margins and uncertainties. Resolving each issue requires increasingly difficult analyses because the aging process has progressively moved the stockpile further from the original test base. While the focus remains on the U.S. nuclear weapons program, where possible, the Program also enables the use of high-performance computing (HPC) and simulation tools to address broader national security needs, such as foreign nuclear weapon assessments and nuclear counterterrorism. The 2022 Nuclear Posture Review (NPR) calls for NNSA to “deliver a modern, adaptive nuclear security enterprise based on an integrated strategy for risk management, production-based resilience, science and technology innovation, and workforce initiatives.” Furthermore, “NNSA will establish a Science and Technology Innovation Initiative to accelerate the integration of science and technology (S&T) throughout its activities.” Executing this strategy necessitates the continued emphasis on developing and sustaining high-quality scientific and engineering staff, as well as supporting computational and experimental capabilities. These components constitute the foundation of the nuclear weapons program. The continued success of the SSP and LEPs is predicated upon the ability to credibly certify the stockpile, without a return to underground nuclear tests (UGTs). Shortly after the nuclear test moratorium entered into force in 1992, the Accelerated Strategic Computing Initiative (ASCI) was established to provide an extensive simulation capability to underpin stockpile certification. While computing and simulation have always been essential to the success of the nuclear weapons program, the program goal of ASCI was to execute NNSA’s vision of using these tools in support of the stockpile stewardship mission. The ASCI program was essential to the successful demonstration of the SSP, providing critical nuclear weapons simulation and modeling capabilities. ASCI officially evolved into the ASC program in fiscal year (FY) 2005, but the mission remains essentially the same: provide the simulation and computational capabilities that underpin the ability to maintain a safe, secure, effective nuclear weapon stockpile, without returning to underground nuclear testing. The capabilities that the ASC program provides at the national laboratories play a vital role in the nuclear security enterprise and are necessary for fulfilling the stockpile stewardship and life extension requirements outlined for NNSA. The Program develops modern simulation tools that provide insights into stockpile aging issues, provide the computational and simulation tools that enable designers and analysts to certify the current stockpile and life-extended nuclear weapons, and inform the decision-making process when any modifications in nuclear warheads or the associated manufacturing processes are deemed necessary. Furthermore, ASC is enhancing the predictive simulation capabilities that are essential to evaluate weapons effects, design experiments, and ensure test readiness. The ASC program continues to improve its unique tools to solve stockpile problems— with a focus on sufficient resolution, dimensionality, and scientific detail—to enable Quantification of Margins and Uncertainties (QMU) and to resolve the increasingly difficult analyses needed for stockpile stewardship. The needs of the Stockpile Management and Production Modernization programs (formerly Directed Stockpile Work) also drive the requirements for simulation and computational resources. These requirements include planned LEPs, stockpile support activities, and mitigation efforts against the potential for technical surprise. All of the weapons within the current stockpile are in some stage of the life extension process. The simulation and computational capabilities are crucial for successful execution of these life extensions and for ensuring NNSA can certify these life-extended weapons without conducting a UGT.

97 MATHEMATICS AND COMPUTING↗

Report on the ASCR Workshop on Modeling and Simulation of Exascale Systems and Applications

In the last few years, numerous studies commissioned by the Department of Energy (DOE), the Defense Advanced Research Projects Agency (DARPA), and others have examined the challenges that will have to be overcome for computational science to achieve exascale, a two order-of-magnitude increase in capability beyond what is possible today. Exponential growth in the number of devices (Moore’s Law) is expected to continue unabated, but the end of Dennard scaling, and hence the growth in performance of individual processors, has led to exponential growth in concurrency. Power consumption has also grown to the point where the cost of the electrical power for systems threatens to exceed the cost of acquiring them. Ever-shrinking devices, and the growth in their number, are expected to increase the rate of soft, or transient, errors such that applications will no longer be able to assume correct behavior of the underlying machine. Adapting to such systems will require new mathematical algorithms that minimize synchronization and data movement, and a new generation of scientific software. Rather than addressing these issues in isolation, a new process of “Co-Design” is being pursued in which application and computer scientists work toward the common goal of an exascale ecosystem of systems and applications. Modeling and simulation (ModSim) is a critical part of this process. It enables scientists and engineers to analyze future algorithms, applications, and computing systems, long before they are realized, and make the necessary design decisions such that exascale science is broadly achieved. Later, when exascale applications and systems are realized, ModSim technology will enable us to understand their behavior, debugging and optimizing them. Given the increasing importance of modeling and simulation, a workshop was organized by DOE/ASCR to assess the current state of the art and to identify research challenges that must be overcome in this field. Fifty experts, representing academia, industry, and government, participated in the workshop, and others contributed both beforehand and afterwards. This report summarizes their findings.

97 MATHEMATICS AND COMPUTING↗

Report on the ASCR Workshop on Modeling and Simulation of Exascale Systems and Applications

In the last few years, numerous studies commissioned by the Department of Energy (DOE), the Defense Advanced Research Projects Agency (DARPA), and others have examined the challenges that will have to be overcome for computational science to achieve exascale, a two order-of-magnitude increase in capability beyond what is possible today. Exponential growth in the number of devices (Moore's Law) is expected to continue unabated, but the end of Dennard scaling, and hence the growth in performance of individual processors, has led to exponential growth in concurrency. Power consumption has also grown to the point where the cost of the electrical power for systems threatens to exceed the cost of acquiring them. Ever-shrinking devices, and the growth in their number, are expected to increase the rate of soft, or transient, errors such that applications will no longer be able to assume correct behavior of the underlying machine. Adapting to such systems will require new mathematical algorithms that minimize synchronization and data movement, and a new generation of scientific software. Rather than addressing these issues in isolation, a new process of "Co-Design" is being pursued in which application and computer scientists work toward the common goal of an exascale ecosystem of systems and applications. Modeling and simulation (ModSim) is a critical part of this process. It enables scientists and engineers to analyze future algorithms, applications, and computing systems, long before they are realized, and make the necessary design decisions such that exascale science is broadly achieved. Later, when exascale applications and systems are realized, ModSim technology will enable us to understand their behavior, debugging and optimizing them. Given the increasing importance of modeling and simulation, a workshop was organized by DOE/ASCR to assess the current state of the art and to identify research challenges that must be overcome in this field. Fifty experts, representing academia, industry, and government, participated in the workshop, and others contributed both beforehand and afterwards. This report summarizes their findings.

97 MATHEMATICS AND COMPUTING↗

Extreme Compression for Large Scale Data Store

For the last 5 years Accelogic pioneered and perfected a radically new theory of numerical computing codenamed “Compressive Computing”, which has an extremely profound impact on real-world computer science [1]. At the core of this new theory is the discovery of one of its fundamental theorems which states that, under very general conditions, the vast majority (typically between 70% and 80%) of the bits used in modern large-scale numerical computations are absolutely irrelevant for the accuracy of the end result. This theory of Compressive Computing provides mechanisms able to identify (with high intelligence and surgical accuracy) the number of bits (i.e., the precision) that can be used to represent numbers without affecting the substance of the end results, as they are computed and vary in real time. The bottom line outcome would be to provide a state-of-the-art compression algorithm that surpasses those currently available in the ROOT framework, with the purpose of enabling substantial economic and operational gains (including speedup) for High Energy and Nuclear Physics data storage/analysis. In our initial studies, a factor of nearly x4 (3.9) compression was achieved with RHIC/STAR data where ROOT compression managed only x1.4.In this contribution, we will present our concepts of “functionally lossless compression”, have a glance at examples and achievements in other communities, present the results and outcome of our current, ongoing R&D, as well as present a high-level view of our plan to move forward with a ROOT implementation that would deliver a basic solution readily integrated into HENP applications. As a collaboration of experimental scientists, private industry, and the ROOT Team, our aim is to capitalize on the substantial success delivered by the initial effort and produce a robust technology properly packaged as an open-source tool that could be used by virtually every experiment around the world as means for improving data management and accessibility.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Progress toward Accelogic compression in ROOT

For the last 7 years, Accelogic pioneered and perfected a radically new theory of numerical computing codenamed “Compressive Computing”, which has an extremely profound impact on real-world computer science [1]. At the core of this new theory is the discovery of one of its fundamental theorems which states that, under very general conditions, the vast majority (typically between 70% and 80%) of the bits used in modern large-scale numerical computations are absolutely irrelevant for the accuracy of the end result. This theory of Compressive Computing provides mechanisms able to identify (with high intelligence and surgical accuracy) the number of bits (i.e., the precision) that can be used to represent numbers without affecting the substance of the end results, as they are computed and vary in real time. The bottom-line outcome will be to provide state-of-the-art compression algorithms --and accompanying software libraries-- able to surpass the performance of the compression engines currently available in the ROOT [7] framework. The resulting technology has the capability to enable substantial economic and operational gains (including speedup) for High Energy and Nuclear Physics data storage/analysis. In our initial studies, a factor of nearly x4 (3.9) compression was achieved with RHIC/STAR data where ROOT compression managed only x1.4 [6].As a collaboration of experimental scientists, private industry, and the ROOT Team, our aim is to capitalize on the substantial success delivered by the initial effort and produce a robust technology properly packaged as an open-source tool that could be used by virtually every experiment around the world as means for improving data management and accessibility.In this contribution, we will present our efforts integrating our concepts of “functionally lossless compression” within the ROOT framework implementation, with the purpose of producing a basic solution readily integrated into HENP applications. We will also present our progress applying this compression through realistic examples of analysis from both the STAR and CMS experiments.

Canal, Ph.↗

Speeding up and reducing memory usage for scientific machine learning via mixed precision

Scientific machine learning (SciML) has emerged as a versatile approach to address complex computational science and engineering problems. Within this field, physics-informed neural networks (PINNs) and deep operator networks (DeepONets) stand out as the leading techniques for solving partial differential equations by incorporating both physical equations and experimental data. However, training PINNs and DeepONets require significant computational resources, including long computational times and large amounts of memory. In search of computational efficiency, training neural networks using half precision (float16) rather than the conventional single (float32) or double (float64) precision has gained substantial interest, given the inherent benefits of reduced computational time and memory consumed. However, we find that float16 cannot be applied to SciML methods, because of gradient divergence at the start of training, weight updates going to zero, and the inability to converge to a local minima. To overcome these limitations, we explore mixed precision, which is an approach that combines the float16 and float32 numerical formats to reduce memory usage and increase computational speed. Our experiments showcase that mixed precision training not only substantially decreases training times and memory demands but also maintains model accuracy. Here, we also reinforce our empirical observations with a theoretical analysis. The research has broad implications for SciML in various computational applications.

97 MATHEMATICS AND COMPUTING↗