Search NASA⌕ Search

SEARCH · Search NASA

Results for “Software Reliability Analysis”

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 217 records · Page 12

Software for Transformative Remedial Action Scheme Tool (TRAST)

The transformative remedial action scheme tool (TRAST) can be applied to improve and validate the power system remedial action scheme (RAS), and further improve the performance of power system operation and control. This tool provides a full suite of advanced functionalities, which are given as follows: 1. Advanced statistical data analysis; 2. OPF-based automated power flow case generation; 3. Customized dynamic simulation in HPC/cloud platform; 4. Machine learning based RAS coefficient prediction; 5. A reliable RAS validation strategy in multiple commercial platforms.

Fan, Xiaoyuan↗

Coupled Multiphysics Modeling of Lithium-Ion Batteries for Automotive Crashworthiness Applications

Considerable advances have been made in battery safety models, but achieving predictive accuracy across a wide range of conditions continues to be challenging. Interactions between dynamically evolving mechanical, electrical, and thermal state variables make model prediction difficult during mechanical abuse scenarios. In this study, we develop a physics-based modeling approach that allows for choosing between different mechanical and electrochemical models depending on the required level of analysis. We demonstrate the use of this approach to connect cell-level abuse response to electrode-level and particle-level transport phenomena. A pseudo-two-dimensional model and simplified single-particle models are calibrated to electrical-thermal cycling data and applied to mechanically induced short-circuit scenarios to understand how the choice of electrochemical model affects the model prediction under abuse scenarios. These models are implemented using user-defined subroutines on ls-dyna finite element software and can be coupled with existing automotive crash safety models.

analysis and design of components↗

Demonstration of a Novel Synchrophasor-based Situational Awareness System: Wide Area Power System Visualization, On-line Event Replay and Early Warning of Grid Problems

Since the large North Eastern power system blackout on August 14, 2003, U.S. electric utilities have spent lot of effort on preventing power system cascading outages. Two of the main causes of the August 14, 2003 blackout were inadequate situational awareness and inadequate operator training In addition to the enhancements of the infrastructure of the interconnected power systems, more research and development of advanced power system applications are required for improving the wide-area security monitoring, operation and planning in order to prevent large- scale cascading outages of interconnected power systems. It is critically important for improving the wide-area situation awareness of the operators or operational engineers and regional reliability coordinators of large interconnected systems. With the installation of large number of phasor measurement units (PMU) and the related communication infrastructure, it will be possible to improve the operators’ situation awareness and to quickly identify the sequence of events during a large system disturbance for the post-event analysis using the real-time or historical synchrophasor data. The purpose of this project was to develop and demonstrate a novel synchrophasor-based comprehensive situational awareness system for control centers of power transmission systems. The developed system named WASA (Wide Area Situation Awareness) is intended to improve situational awareness at control centers of the power system operators and regional reliability coordinators. It consists of following main software modules: • Wide-area visualizations of real-time frequency, voltage, and phase angle measurements and their contour displays for security monitoring. • Online detection and location of a major event (location, time, size, and type, such as generator or line outage). • Near-real-time event replay (in seconds) after a major event occurs. • Early warning of potential wide-area stability problems. The system has been deployed and demonstrated at the Tennessee Valley Authority (TVA) and ISO New England system using real-time synchrophasor data from openPDC. Apart from the software product, the outcome of this project consists of a set of technical reports and papers describing the mathematical foundations and computational approaches of different tools and modules, implementation issues and considerations, lessons learned, and the results of lidation processes.

24 POWER TRANSMISSION AND DISTRIBUTION↗

Design Methods, Tools, and Data for Ceramic Solar Receivers

This report presents the development of tools and methods for evaluating the reliability and performance of ceramic materials in high temperature solar receivers. As Concentrating Solar Power (CSP) technologies aim for higher operating temperatures to enhance efficiency and meet industrial process heat requirements, current high temperature metallic materials face challenges in maintaining structural integrity. This report explores advanced ceramics as a promising alternative, given their superior high temperature strength and lower thermal expansion, compared to metals. To address the need for effective ceramic receiver design tools, this report integrates statistical failure models of ceramics into the existing srlife tool: an open-source software package designed to estimate the life of high temperature CSP components. These failure models account for the inherent variability and flaw distribution in ceramics, as well as the impact of subcritical crack growth under high temperature cyclic loads. The report also presents experimental data collected for a commercially available ceramic material, SiC, and details the process of estimating reliability model parameters from these data. A comparative design analysis is then performed between ceramic (SiC) and metallic (current nickel-based superalloys A740H and A282) receiver. This comparison demonstrates that SiC receivers can achieve service life exceeding 30 years under high incident heat flux conditions, compared to just a few years for metallic receivers.

14 SOLAR ENERGY↗

Characterization of Infrasonic Signatures of Earth-Grazing Fireballs as Analogues to Hypersonic Vehicles (Final Report)

Accurate detection, discrimination, and characterization of high-altitude hypersonic events using infrasonic monitoring are critical to planetary defense and global strategic surveillance. This report synthesizes recent advances achieved through rigorous analysis of infrasonic signatures from natural meteoroids, emphasizing shallow entry-angle meteoroids as essentially proxies for artificial hypersonic systems. Meteoroids naturally encompass diverse velocities, trajectories, altitudes, and fragmentation behaviors, enabling systematic validation of empirical period–yield relationships, waveform morphology classifiers, and trajectory-induced back-azimuth deviation models. Integration of adaptive array-processing enhancements within Cardinal software further extends infrasonic detection sensitivity and signal classification reliability. Collectively these advances, based solely on infrasonic signatures or limited optical data, offer robust methodologies for distinguishing natural from artificial hypersonic sources, significantly reducing event geolocation uncertainties and refining source-function determination. The outcomes detailed herein lay foundational groundwork for improved global hypersonic event-surveillance frameworks, supporting improved security preparedness and informing strategic monitoring and defense policies.

54 ENVIRONMENTAL SCIENCES↗

Dynamic Building Load Control to Facilitate High Penetration of Solar Photovoltaic Generation (Final Technical Report)

Solar photovoltaic (PV) resources are the most common form of distributed generation in residential and commercial customer premises within electric distribution networks. A higher penetration of PV generation in distribution circuits will impose challenges on maintaining service voltages within the range of industry standards, power quality, and power flow. Buildings consume 74% of the electricity produced in the United States, and a significant portion of the building load is dispatchable, making them responsive to electrical grid needs. Oak Ridge National Laboratory—in collaboration with Southern Company; the University of Tennessee, Knoxville; and the Georgia Institute of Technology—is examining the PV integration issues in distribution-level electrical grids and developing integrated demand-side control and communication systems to enable responsive loads. The proposed responsive loads mechanism performs renewable generation following to increase the penetration of solar PV within each feeder. The specific objectives of this project are to (1) examine distribution-level PV integration scenarios to understand requirements, (2) undertake an end-to-end simulation-based design of a distributed control strategy of loads geographically near the PV generation asset to minimize the effect on the distribution feeder, (3) deploy and demonstrate the control technology developed in partnership with utilities, and (4) perform a scalability analysis at the utility scale. This 3-year integrated project aims to develop, demonstrate, and validate demand-side control technology to enable increased the penetration of renewables while mitigating challenges that arise due to their intermittency. Activities in Budget Period (BP) 1 focused on a literature review and the formal design of a control system for integrating local distribution with generation and loads. The team used modeling and simulation to evaluate the impact of varying buildings loads, variable PV generation, and power flow dynamics on the distribution circuit. The dynamic models developed in BP 1 were used in BP 2 to develop a model-based control design and a test bed. The test bed has enabled the simulation-based testing and comparison of different control designs and formulations applied to different configurations of the distribution grid, PVs, and building loads. The control approaches developed in BP 2 were implemented in BP 3 in the form of hardware deployed at the Central Baptist Church (CBC) in Knoxville, Tennessee, for testing and evaluation. The outcome of this project was the development and demonstration of open-source, low-cost, low-touch sensing and control retrofits to distributed PV generation and building loads that, in a coordinated fashion, provide the load-shaping response needed to integrate high levels of renewable penetration. This research addresses the target metrics by dynamically controlling a load with solar generation variability to minimize the extent of two-way power flow, enhance reliability, facilitate high PV penetration (>100% of peak load in a line segment), and generate scalable software and hardware solutions adaptable to any penetration levels. The research and development activities are focused and designed to be impactful within the relevant 2020 targets time frame.An accurate open-source integration simulation framework for end-to-end control design was developed and deployed at the CBC facility for testing and evaluation. This final report provides a detailed review of the technical results achieved during this 3-year integrated project. A novel spectral analysis of PV data is demonstrated to derive the requirements of the control design. A detailed simulation-based analysis of PV integration at increasing penetration levels is presented using 1 year of PV data to demonstrate the impact on the distribution circuits. Two different control strategies were developed and demonstrated via simulation to track variable PV generation with adaptive load dispatch. The report concludes with a summary of accomplishments and recommendations for a path forward.

14 SOLAR ENERGY↗

NRC technical basis for evaluation of its position on protection against common cause failure in digital systems used in nuclear power plants - 378

Digital technology has advantages over analog systems, including automated monitoring and alerts for standby safety functions, and predictive algorithms to maintain critical safety systems. Additionally, digital technology generally has higher reliability and can be designed to reduce single point vulnerabilities. For these reasons many nuclear plants have applied digital technology to safety and non-safety related applications, including reactor protection system, feedwater and turbine controls, etc. with a corresponding significant improvement in trip reduction. Nonetheless, digital instrumentation and control (I and C) systems also present potential new vulnerabilities that need to be assessed, including potential failures due to increased complexity of digital systems, the introduction of unique failure modes due to software (including software common cause failure (CCF)), and limited operating history of digital systems in nuclear safety related applications compared to analog systems. The fact that software is intangible means that common methods, such as analysis or testing, used for detecting CCF may not be effective when applied to software. Consequently, digital technology is perceived to pose a potential risk from the introduction of undetected systematic faults that could result in CCF. Despite the I and C system upgrades and modifications performed to date, the U.S. Nuclear Regulatory Commission (NRC) and industry stakeholders have identified the need to modernize the regulatory infrastructure to efficiently address risks associated with the use of digital technology for nuclear safety applications and address regulatory uncertainties. The NRC's current position on CCF is guided by the staff requirements memorandum (SRM) on SECY 93-087. The SRM provides specific acceptance criteria for the evaluation of CCF, which the staff implemented in the Branch Technical Position (BTP) 7-19. However, industry stakeholders have proposed using methods to characterize the likelihood of software CCF and eliminate it from further consideration in a defense-in-depth and diversity analysis. The NRC's current position does not consider these alternatives, and thus corresponding acceptance criteria is not currently available. The work discussed in this paper assesses the underlying technical basis associated with CCF, provides technical support for updating the NRC position and considers proposed methods for addressing potential CCF in digital systems while enhancing efficiency, clarity, and confidence. (authors)

22 GENERAL STUDIES OF NUCLEAR REACTORS↗

Best-estimate plus uncertainty thermal-hydraulic stability analysis of BWRs using TRACG code

Over the last decade, Boiling Water Reactor (BWR) power up-rates have increased plant rated power output significantly. Subsequent projects have expanded flow domains (e.g. MELLLA+) for operation at these higher power levels. This has resulted in an increase in the power to flow ratio in regions susceptible to reactor thermal-hydraulic instabilities. Since BWRs are susceptible to coupled thermal-hydraulic/nuclear oscillations when operating at these conditions, such oscillations must be prevented or reliably detected and suppressed. The Detect and Suppress Solution - Confirmation Density (DSS-CD) is the most sophisticated GEH BWR instability protection system ever employed. DSS-CD implements algorithms that monitor closely-spaced groups of Local Power Range Monitor (LPRM) detectors to detect periodic behavior typical of reactor instability events. This system is able to detect small, localized power variations in the core, distinguish between true instabilities and plant noise, and trip/scram the reactor while maintaining adequate safety margins. The combination of hardware, software, and system setpoints provides protection against violation of the Safety Limit Minimum Critical Power Ratio (SLMCPR) for anticipated oscillations. To support DSS-CD implementation, the TRACG system code is used to simulate events to confirm the capability of the DSS-CD solution for early oscillation detection and suppression. TRACG is a GEH proprietary version of the Transient Reactor Analysis Code (TRAC). TRACG includes a multi-dimensional, two-fluid model for the reactor thermal-hydraulics and a three-dimensional reactor kinetics model. The models are qualified to simulate a large variety of tests and reactor configurations, including thermal-hydraulic stability events. These features allow for detailed, best-estimate simulation of a wide range of BWR phenomena. A set of integrated TRACG event simulations for reasonably limiting anticipated events can be used to calculate the effect on the Minimum Critical Power Ratio (MCPR) performance. The purpose of the DSS-CD TRACG analysis is to confirm the inherent MCPR margin afforded by the solution design. This paper presents the Best Estimate Plus Uncertainty (BEPU) DSS-CD TRACG methodology and its application to BWR Thermal-Hydraulic (T-H) stability analyses. The statistical Code Scaling, Applicability and Uncertainty (CSAU) methodology (defined in NUREG/CR-5249) is used to calculate the MCPR uncertainty. The TRACG simulation includes a full core individual bundle model in which each fuel bundle is modeled as an individual T-H channel. The complete CSAU analysis of full core individual bundle model is an innovative solution represents the state-of-the-art stability analysis of BWRs and is the first ever full statistical analysis for stability safety analyses. The adoption of BEPU methodologies for stability analyses advances the understanding of the associated physical phenomena and maintains the safety of reactor plant operation in expanded operation domain with up-rated power. (authors)

21 SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLANTS↗

SMATool: Strength of materials analysis toolkit

The study of the strength of materials is a cornerstone in material science and engineering, playing a critical role in shaping the progress and application of materials in diverse industrial sectors. The strength of a material is meticulously examined to understand the behavior of the material under different stress conditions and environments, thereby guiding material selection and structural design. Herein, we introduce the SMATool, a computational toolkit for the efficient calculation and analysis of material strength at both zero and finite temperatures for 3D, 2D, 1D, and tubular 2D-based nanostructures and nanotubes, as well as 1D nanoribbons. The toolkit is capable of calculating tensile, shear, ultimate, yield, and indentation (Vickers' hardness) strengths in various dimensions, as well as the energy storage capacity. We conducted several calculations both at zero and finite temperatures to validate the accuracy and reliability of the developed software. Here, the results show that the SMATool package provides accurate predictions that align with existing data on material strength. SMATool integrates seamlessly with widely used electronic structure codes like VASP and Quantum Espresso, providing a user-friendly interface catering to academic researchers and industry professionals.

36 MATERIALS SCIENCE↗

Build and Execute Environment

At exascale, the challenge becomes to develop applications that run at scale and use exascale platforms reliably, efficiently, and flexibly. Workflows become much more complex because they must seamlessly integrate simulation and data analytics. They must include down-sampling, post-processing, feature extraction, and visualization. Power and data transfer limitations require these analysis tasks to be run in-situ or in-transit. We expect successful workflows will comprise multiple linked simulations along with tens of analysis routines. Users will have limited development time at scale and, therefore, must have rich tools to develop, debug, test, and deploy applications. At this scale, successful workflows will compose linked computations from an assortment of reliable, well-defined computation elements, ones that can come and go as required, based on the needs of the workflow over time. We propose a novel framework that utilizes both virtual machines (VMs) and software containers to create a workflow system that establishes a uniform build and execution environment (BEE) beyond the capabilities of current systems. In this environment, applications will run reliably and repeatably across heterogeneous hardware and software. Containers, both commercial (Docker and Rocket) and open-source (LXC and LXD), define a runtime that isolates all software dependencies from the machine operating system. Workflows may contain multiple containers that run different operating systems, different software, and even different versions of the same software. We will run containers in open-source virtual machines (KVM) and emulators (QEMU) so that workflows run on any machine entirely in user-space. On this platform of containers and virtual machines, we will deliver workflow software that provides services, including repeatable execution, provenance, checkpointing, and future proofing. We will capture provenance about how containers were launched and how they interact to annotate workflows for repeatable and partial re-execution. We will coordinate the physical snapshots of virtual machines with parallel programming constructs, such as barriers, to automate checkpoint and restart. We will also integrate with HPC-specific container runtimes to gain access to accelerators and other specialized hardware to preserve native performance. Containers will link development to continuous integration. When application developers check code in, it will automatically be tested on a suite of different software and hardware architectures.

Guan, Qiang↗

Accelerated Data Analytics for Power-System Time-Series (ADAPT) [Final Project Report]

Over the past decade, electric utilities have made significant progress in deploying networks of phasor measurement units (PMUs). The high-speed measurements provided by PMUs are valuable for offline analysis, but the sheer volume of data presents a challenge for utilities. The Accelerated Data Analytics for Power-System Time-Series (ADAPT) project was created to address this challenge by developing SciSync, an open-source software tool that enables utilities to rapidly read, process, analyze, and review grid measurements. SciSync was designed to provide the capabilities of Archive Walker, a research tool developed by PNNL, along with the reliability and performance of commercial-grade software.

24 POWER TRANSMISSION AND DISTRIBUTION↗

Build and Execute Environment

At exascale, the challenge becomes to develop applications that run at scale and use exascale platforms reliably, efficiently, and flexibly. Workflows become much more complex because they must seamlessly integrate simulation and data analytics. They must include down-sampling, post-processing, feature extraction, and visualization. Power and data transfer limitations require these analysis tasks to be run in-situ or in-transit. We expect successful workflows will comprise multiple linked simulations along with tens of analysis routines. Users will have limited development time at scale and, therefore, must have rich tools to develop, debug, test, and deploy applications. At this scale, successful workflows will compose linked computations from an assortment of reliable, well-defined computation elements, ones that can come and go as required, based on the needs of the workflow over time. We propose a novel framework that utilizes both virtual machines (VMs) and software containers to create a workflow system that establishes a uniform build and execution environment (BEE) beyond the capabilities of current systems. In this environment, applications will run reliably and repeatably across heterogeneous hardware and software. Containers, both commercial (Docker and Rocket) and open-source (LXC and LXD), define a runtime that isolates all software dependencies from the machine operating system. Workflows may contain multiple containers that run different operating systems, different software, and even different versions of the same software. We will run containers in open-source virtual machines (KVM) and emulators (QEMU) so that workflows run on any machine entirely in user-space. On this platform of containers and virtual machines, we will deliver workflow software that provides services, including repeatable execution, provenance, checkpointing, and future proofing. We will capture provenance about how containers were launched and how they interact to annotate workflows for repeatable and partial re-execution. We will coordinate the physical snapshots of virtual machines with parallel programming constructs, such as barriers, to automate checkpoint and restart. We will also integrate with HPC-specific container runtimes to gain access to accelerators and other specialized hardware to preserve native performance. Containers will link development to continuous integration. When application developers check code in, it will automatically be tested on a suite of different software and hardware architectures.

Guan, Qiang↗

Developing a Vorticity-Velocity-Based Off-Body Solver to Perform Multifidelity Simulations of Wind Farms

Wind power has become a key player in satisfying the global energy needs. With increased market penetration, unanticipated unsteady loading induced failures, installation related reductions in power generation, and significant maintenance costs have underscored the need to predict the unsteady fluid-structure interactions related to turbine layout and off-design wind conditions. Contemporary turbine design tools are incapable of accounting for such loadings. As a result, researchers have started utilizing high-Performance-Computing (HPC) based Computational Fluid Dynamics (CFD) solvers, such as the U.S. Department of Energy sponsored ExaWind software package, to investigate these phenomena. Unfortunately, such HPC tools are computationally expensive for routine industrial use, often because of the sheer number of cells required to resolve the wake flowfield. This paper describes a preliminary effort to address this issue by developing a vorticity-velocity based CFD off-body solver, VorTran-M2-AMReX, that integrates directly with DOE's ExaWind wind turbine analysis system to perform accurate and reliable simulations of wind turbine/farm at a lower computational cost than ExaWind alone. This article summarizes work undertaken to date concerning the assembly of the proposed analysis tool, and provides preliminary validation and verification of the VorTran-M2-AMReX off-body solver.

adaptive mesh refinement↗

A New Modeling Framework for Geothermal Operational Optimization with Machine Learning (GOOML)

Geothermal power plants are excellent resources for providing low carbon electricity generation with high reliability. However, many geothermal power plants could realize significant improvements in operational efficiency from the application of improved modeling software. Increased integration of digital twins into geothermal operations will not only enable engineers to better understand the complex interplay of components in larger systems but will also enable enhanced exploration of the operational space with the recent advances in artificial intelligence (AI) and machine learning (ML) tools. Such innovations in geothermal operational analysis have been deterred by several challenges, most notably, the challenge in applying idealized thermodynamic models to imperfect as-built systems with constant degradation of nominal performance. This paper presents GOOML: a new framework for Geothermal Operational Optimization with Machine Learning. By taking a hybrid data-driven thermodynamics approach, GOOML is able to accurately model the real-world performance characteristics of as-built geothermal systems. Further, GOOML can be readily integrated into the larger AI and ML ecosystem for true state-of-the-art optimization. This modeling framework has already been applied to several geothermal power plants and has provided reasonably accurate results in all cases. Therefore, we expect that the GOOML framework can be applied to any geothermal power plant around the world.

15 GEOTHERMAL ENERGY↗

Integrating PCTRAN with AI-Driven Host-Intrusion Detection and Secured Container Systems for Advanced Malware Analysis (Summer Internship Report)

This study presents a solution for enhancing the security of the Personal Computer Transient Analyzer (PCTRAN) PC-based Nuclear Power Plant Simulator by integrating the software with an artificial intelligence (AI)-driven host-intrusion detection system (HIDS), in addition to a secured container system, for malware analysis. PCTRAN is a Windows XP-based software package that has the ability to simulate a variety of accident and transient conditions for nuclear power plants (NPPs). It offers a high-resolution replica of the Nuclear Steam Supply System (NSSS) and displays the status of important parameters allowing for operator interaction. By including AI-driven HIDS for the NSSS, the framework can identify security threats in real-time, ensuring the integrity of the nuclear simulation environment. Additionally, the secured container system offers the ability to isolate and analyze malware, preventing potential threats from affecting core systems. The integration process involves extensive testing and validation in order to ensure accuracy, reliability, and compliance with security policies. This framework sets a new precedent for secure simulation and training in NPP operations, and offers insight for future advancements in cybersecurity.

21 SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLANTS↗

Using Micro-Synchrophasor Data for Advanced Distribution Grid Planning and Operations Analysis

This report reviews the potential for distribution-grid phase-angle data that will be available from new micro-synchrophasors (µPMUs) to be utilized in existing distribution-grid planning and operations analysis. This data could augment the current diagnostic capabilities of grid analysis software, used in both planning and operations for applications such as fault location, and provide data for more accurate modeling of the distribution system. µPMUs are new distribution-grid sensors that will advance measurement and diagnostic capabilities and provide improved visibility of the distribution grid, enabling analysis of the grid’s increasingly complex loads that include features such as large volumes of distributed generation. Large volumes of DG leads to concerns on continued reliable operation of the grid, due to changing power flow characteristics and active generation, with its own protection and control capabilities. Using µPMU data on change in voltage phase angle between two points in conjunction with new and existing distribution-grid planning and operational tools is expected to enable model validation, state estimation, fault location, and renewable resource/load characterization. Our findings include: data measurement is outstripping the processing capabilities of planning and operational tools; not every tool can visualize a voltage phase-angle measurement to the degree of accuracy measured by advanced sensors, and the degree of accuracy in measurement required for the distribution grid is not defined; solving methods cannot handle the high volumes of data generated by modern sensors, so new models and solving methods (such as graph trace analysis) are needed; standardization of sensor-data communications platforms in planning and applications tools would allow integration of different vendors’ sensors and advanced measurement devices. In addition, data from advanced sources such as µPMUs could be used to validate models to improve/ensure accuracy, providing information on normally estimated values such as underground conductor impedance, and characterization of complex loads. Although the input of high-fidelity data to existing tools will be challenging, µPMU data on phase angle (as well as other data from advanced sensors) will be useful for basic operational decisions that are based on a trend of changing data.

25 ENERGY STORAGE↗

Large Scale Computing and Storage Requirements for Biological and Environmental Research

In May 2009, NERSC, DOE's Office of Advanced Scientific Computing Research (ASCR), and DOE's Office of Biological and Environmental Research (BER) held a workshop to characterize HPC requirements for BER-funded research over the subsequent three to five years. The workshop revealed several key points, in addition to achieving its goal of collecting and characterizing computing requirements. Chief among them: scientific progress in BER-funded research is limited by current allocations of computational resources. Additionally, growth in mission-critical computing -- combined with new requirements for collaborative data manipulation and analysis -- will demand ever increasing computing, storage, network, visualization, reliability and service richness from NERSC. This report expands upon these key points and adds others. It also presents a number of"case studies" as significant representative samples of the needs of science teams within BER. Workshop participants were asked to codify their requirements in this"case study" format, summarizing their science goals, methods of solution, current and 3-5 year computing requirements, and special software and support needs. Participants were also asked to describe their strategy for computing in the highly parallel,"multi-core" environment that is expected to dominate HPC architectures over the next few years.

54 ENVIRONMENTAL SCIENCES↗

Automated Cache Performance Analysis And Optimization

While there is no lack of performance counter tools for coarse-grained measurement of cache activity, there is a critical lack of tools for relating data layout to cache behavior to application performance. Generally, any nontrivial optimizations are either not done at all, or are done ”by hand” requiring significant time and expertise. To the best of our knowledge no tool available to users measures the latency of memory reference instructions for particular addresses and makes this information available to users in an easy-to-use and intuitive way. In this project, we worked to enable the Open|SpeedShop performance analysis tool to gather memory reference latency information for specific instructions and memory addresses, and to gather and display this information in an easy-to-use and intuitive way to aid performance analysts in identifying problematic data structures in their codes. This tool was primarily designed for use in the supercomputer domain as well as grid, cluster, cloud-based parallel e-commerce, and engineering systems and middleware. Ultimately, we envision a tool to automate optimization of application cache layout and utilization in the Open|SpeedShop performance analysis tool. To commercialize this software, we worked to develop core capabilities for gathering enhanced memory usage performance data from applications and create and apply novel methods for automatic data structure layout optimizations, tailoring the overall approach to support existing supercomputer and cluster programming models and constraints. In this Phase I project, we focused on infrastructure necessary to gather performance data and present it in an intuitive way to users. With the advent of enhanced Precise Event-Based Sampling (PEBS) counters on recent Intel processor architectures and equivalent technology on AMD processors, we are now in a position to access memory reference information for particular addresses. Prior to the introduction of PEBS counters, cache behavior could only be measured reliably in the aggregate across tens or hundreds of thousands of instructions. With the newest iteration of PEBS technology, cache events can be tied to a tuple of instruction pointer, target address (for both loads and stores), memory hierarchy, and observed latency. With this information we can now begin asking questions regarding the efficiency of not only regions of code, but how these regions interact with particular data structures and how these interactions evolve over time. In the short term, this information will be vital for performance analysts understanding and optimizing the behavior of their codes for the memory hierarchy. In the future, we can begin to ask how data layouts might be changed to improve performance and, for a particular application, what the theoretical optimal performance might be. The overall benefit to be produced by this effort was a commercial quality easy-to- use and scalable performance tool that will allow both beginner and experienced parallel programmers to automatically tune their applications for optimal cache usage. Effective use of such a tool can literally save weeks of performance tuning effort. Easy to use. With the proposed innovations, finding and fixing memory performance issues would be more automated and hide most to all of the performance engineer expertise ”under the hood” of the Open|SpeedShop performance tool. One of the biggest public benefits from the proposed innovations is that it makes performance analysis more usable to a larger group of application developers. Intuitive reporting of results. The Open|SpeedShop performance analysis tool has a rich set of intuitive, yet detailed reports for presenting performance results to application developers. Our goal was to leverage this existing technology to present the results from our memory performance addition to Open|SpeedShop. Suitable for experts as well as novices. Application performance is getting more difficult to measure as the hardware platforms they run on become more complicated. This makes life difficult for the application developer, in that they need to know more about the hardware platform, including the memory system hierarchy, in order to understand the performance of their application. Some application developers are comfortable in that sceario, while others want to do their scientific research and not have to understand all the nuances in the hardware platform they are running their application on. Our proposed innovations were aimed to support both experts and novice performance analysts. Useful in many markets. The enhancement to Open|SpeedShop would appeal to a broader market space, as it will be useful in scientific, commercial, and cloud computing environments. Our goal was to use technology developed initially at the and Lawrence Livermore National Laboratory combined with the development and commercial software experience of the Argo Navis Technologies, LLC (ANT) to form a powerful combination to delivery these objectives.

97 MATHEMATICS AND COMPUTING↗