Search NASA⌕ Search

SEARCH · Search NASA

Results for “software performance”

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 595 records · Page 33

Accelerating Scientific Applications With SambaNova Reconfigurable Dataflow Architecture

Here, our exploratory work finds that the SambaNova Reconfigurable Dataflow Architecture (RDA) along with the SambaFlow software stack provides for an attractive system and solution to accelerate AI for science workloads. We have observed the efficacy of using the system with a diverse set of science applications and reasoned their suitability for performance gains over traditional hardware. As the Data-Scale system provides for a very large memory capacity, the system can be used to train models that typically do not fit in a GPU. The architecture also provides for deeper integration with upcoming supercomputers at the Argonne Leadership Computing Facility (ALCF), a US Department of Energy Office of Science user facility, to help advance science insights.

97 MATHEMATICS AND COMPUTING↗

Trilinos: Enabling Scientific Computing across Diverse Hardware Architectures at Scale

Trilinos is a community-developed, open-source software framework that facilitates building large-scale, complex, multiscale, multiphysics simulation code bases for scientific and engineering problems. Since the Trilinos framework has undergone substantial changes to support new applications and new hardware architectures, this document is an update to “An Overview of the Trilinos project” by Heroux et al. (ACM Transactions on Mathematical Software, 31(3):397–423, 2005). It describes the design of Trilinos, introduces its new organization in product areas, and highlights established and new features available in Trilinos. Particular focus is put on the modernized software stack based on the Kokkos ecosystem to deliver performance portability across heterogeneous hardware architectures. This article also outlines the organization of the Trilinos community and the contribution model to help onboard interested users and contributors.

Heterogeneous Hardware Architectures↗

ORCHA: A performance portability system for extreme heterogeneity

Heterogeneity is the prevalent trend in the rapidly evolving high-performance computing (HPC) landscape in both hardware and application software. The diversity in hardware platforms, currently comprising various accelerators and a future possibility of specializable chiplets, poses a significant challenge for scientific software developers aiming to harness optimal performance across different computing platforms while maintaining the quality of solutions when their applications are simultaneously growing more complex. Code synthesis and code generation can provide mechanisms to mitigate this challenge. We have developed a divide and conquer approach where different aspects of performance are handled by different stand-alone tools that are interfaced with the application through generated code. This portability system, ORCHA, enables users to configure and orchestrate their computations among available resources on a platform by specifying a high-level recipe, thereby permitting a many-to-many paradigm where each recipe results in a different variant of the application. The core design goal is to let users decide the application’s hardware mapping and orchestration by editing only the high-level recipe—without modifying the maintained source code or binding the application to a particular runtime system. Tools in ORCHA distribution are: CG-Kit for translating the recipe into an execution graph; Milhoja to execute the graph by orchestrating data and task movement among hardware resources; and Macroprocessor that enables users to define their own code-shorthand for higher composability and easier management of code variants. Additionally, the design of ORCHA permits tools to work in a plug-and-play mode where the application can build and run without CG-Kit and Milhoja, and either tool can be swapped out for other tools with similar capabilities by modifying the code generation portion of ORCHA. In this paper, we describe the design of ORCHA and the role that code-generation plays in isolating applications from tools. We demonstrate the breadth of configurations ORCHA enables with a case study in which an application configuration is realized on three distinct hardware mappings—a GPU-centric, a CPU/GPU balanced, and a CPU/GPU concurrent layouts by using different recipes.

Lee, Youngjun↗

Extreme-Scale Algorithms & Software Resilience (EASIR) Architecture-Aware Algorithms for Scalable Performance and Resilience on Heterogeneous Architectures

This project addresses both communication-avoiding algorithms, and reproducible floating-point computation. Communication, i.e. moving data, either between levels of memory or processors over a network, is much more expensive per operation than arithmetic (measured in time or energy), so we seek algorithms that greatly reduce communication. We developed many new algorithms for both dense and sparse, and both direct and iterative linear algebra, attaining new communication lower bounds, and getting large speedups in many cases. We also extended this work in several ways: (1) We minimize writes separately from reads, since writes may be much more expensive than reads on emerging memory technologies, like Flash, sometimes doing asymptotically fewer writes than reads. (2) We extend the lower bounds and optimal algorithms to arbitrary algorithms that may be expressed as perfectly nested loops accessing arrays, where the array subscripts may be arbitrary affine functions of the loop indices (eg A(i), B(i,j+k, k+3*m-7, …) etc.). (3) We extend our communication-avoiding approach to some machine learning algorithms, such as support vector machines. This work has won a number of awards. We also address reproducible floating-point computation. We define reproducibility to mean getting bitwise identical results from multiple runs of the same program, perhaps with different hardware resources or other changes that should ideally not change the answer. Many users depend on reproducibility for debugging or correctness. However, dynamic scheduling of parallel computing resources, combined with nonassociativity of floating point addition, makes attaining reproducibility a challenge even for simple operations like summing a vector of numbers, or more complicated operations like the Basic Linear Algebra Subprograms (BLAS). We describe an algorithm that computes a reproducible sum of floating point numbers, independent of the order of summation. The algorithm depends only on a subset of the IEEE Floating Point Standard 754-2008, uses just 6 words to represent a “reproducible accumulator,” and requires just one read-only pass over the data, or one reduction in parallel. New instructions based on this work are being considered for inclusion in the future IEEE 754-2018 floating-point standard, and new reproducible BLAS are being considered for the next version of the BLAS standard.

97 MATHEMATICS AND COMPUTING↗

Benchmarking of Software and Methods for Use in Transient Multidimensional Fuel Performance with Spatial Reactor Kinetics

The key physics involved in accurate prediction of reactor-fuel-element behavior includes neutron transport and thermal hydraulics. The thermal hydraulic feedback mechanism is primarily provided through cross sections to the neutron transport that are temperature and density dependent. Historically, this coupling was primarily seen only in reactor simulators, which are well suited to model the reactor core, giving only a coarse treatment to individual fuel pins as well as simple models for thermal distribution calculations. This poor resolution on the primary coupling mechanisms can lead to conservatisms that should be removed to improve fuel design and performance. This work seeks to address the resolution of space-time-dependent neutron kinetics with thermal feedback within the fuel pin scale in the multiphysics framework. The specific application of this new capability is transient performance analysis of space-time-dependent temperature distribution of fuel elements. The coupling between the neutron transport and the thermal feedback is extremely important in this highly coupled problem, primarily applicable to reactivity-initiated-accidents (RIAs) and loss-of-coolant-accidents (LOCAs). The capability developed will include the coupling of the time-dependent neutron transport with the time-dependent thermal diffusion capability. An improvement in resolution and coupling is proposed by developing neutron transport models that are internally coupled with high fidelity within fuel pin thermal calculations in a multiphysics framework. Good agreement is shown with benchmarks and problems from the literature of RIAs and LOCAs for the tools used.

22 GENERAL STUDIES OF NUCLEAR REACTORS↗

Benchmarking of software and methods for use in transient multidimensional fuel performance with spatial reactor kinetics

The key physics involved in accurate prediction of reactor-fuel-element behavior includes neutron transport and thermal hydraulics. The thermal hydraulic feedback mechanism is primarily provided through cross sections to the neutron transport that are temperature and density dependent. Historically, this coupling was primarily seen only in reactor simulators, which are well suited to model the reactor core, giving only a coarse treatment to individual fuel pins as well as simple models for thermal distribution calculations. This poor resolution on the primary coupling mechanisms can lead to conservatisms that should be removed to improve fuel design and performance. This work seeks to address the resolution of space-time-dependent neutron kinetics with thermal feedback within the fuel pin scale in the multi-physics framework. The specific application of this new capability is transient performance analysis of space-time-dependent temperature distribution of fuel elements. The coupling between the neutron transport and the thermal feedback is extremely important in this highly coupled problem, primarily applicable to reactivity-initiated- accidents (RIAs) and loss-of-coolant-accidents (LOCAs). The capability developed will include the coupling of the time-dependent neutron transport with the time-dependent thermal diffusion capability. An improvement in resolution and coupling is proposed by developing neutron transport models that are internally coupled with high fidelity within fuel pin thermal calculations in a multi-physics framework. Good agreement is shown with benchmarks and problems from the literature of RIAs and LOCAs for the tools used. (authors)

22 GENERAL STUDIES OF NUCLEAR REACTORS↗

Model Predictive Control of Integrated Gasification Combined Cycle Power Plants

The primary project objectives were to understand how the process design of an integrated gasification combined cycle (IGCC) power plant affects the dynamic operability and controllability of the process. Steady-state and dynamic simulation models were developed to predict the process behavior during typical transients that occur in plant operation. Advanced control strategies were developed to improve the ability of the process to follow changes in the power load demand, and to improve performance during transitions between power levels. Another objective of the proposed work was to educate graduate and undergraduate students in the application of process systems and control to coal technology. Educational materials were developed for use in engineering courses to further broaden this exposure to many students. ASPENTECH software was used to perform steady-state and dynamic simulations of an IGCC power plant. Linear systems analysis techniques were used to assess the steady-state and dynamic operability of the power plant under various plant operating conditions. Model predictive control (MPC) strategies were developed to improve the dynamic operation of the power plants. MATLAB and SIMULINK software were used for systems analysis and control system design, and the SIMULINK functionality in ASPEN DYNAMICS was used to test the control strategies on the simulated process. Project funds were used to support a Ph.D. student to receive education and training in coal technology and the application of modeling and simulation techniques.

01 COAL, LIGNITE, AND PEAT↗

Development of a CPU/GPU portable software library for Lagrangian–Eulerian simulations of liquid sprays

The Lagrangian–Eulerian method is widely used for simulations of fuel sprays in turbulent combustion because of the advantage of treating the spray droplets as discrete points. One challenge of the Lagrangian–Eulerian method is the intense computational requirement when tracking the large number of Lagrangian particles needed for high fidelity. We have developed a performance-portable library, Grit, to track the Lagrangian particles in parallel on central processing unit (CPU) and graphics processing unit (GPU) accelerated high performance computing (HPC) architectures. Grit is a C++ library which employs Message Passing Interface (MPI) for distributed memory parallelism and Kokkos programming model for on-node shared memory parallelism with performance portability across different architectures of GPUs and multi-core/manycore CPUs. The parallel algorithms, key parallel kernels, and their performances on the pre-exascale supercomputer, Summit, are presented. Grit is coupled with a direct numerical simulation (DNS) solver, S3D, for multiphase simulations. A conservative formulation has been developed and implemented in Grit for phase coupling with thermodynamic consistency. The formulation separates the conservation of mass, momentum and energy from the physical models to prevent accidental violation of conservation laws due to inconsistent models. The formulation also enforces consistent definitions of enthalpies of the fuel for both phases and the latent heat of evaporation. Finally, simulations of turbulent particle-laden flow and the evaporation of dilute turbulent spray jet are performed to verify the software implementation, and to demonstrate the scalability of Grit for large-scale multiphase simulations.

97 MATHEMATICS AND COMPUTING↗

Simulation and study of the influence of the buffer intrinsic layer, back-surface field, densities of interface defects, resistivity of p-type silicon substrate and transparent conductive oxide on heterojunction with intrinsic thin-layer (HIT) solar cell

The influence of various parameters such as buffer intrinsic layers, back-surface fields, densities of interface defects (D{sub it}), the resistivity of p-type silicon substrates ({rho}) and then work function of transparent conductive oxide ({phi}{sub TCO}) on heterojunction with intrinsic thin-layer (HIT) solar cell performance was investigated using software simulation. Automat for the simulation of heterostructures (AFORS-HET) software was used for that purpose. Our results indicate that band bending, which is determined by the band offsets at the buffer intrinsic/c-Si and/or the c-Si/back-surface field heterointerface, could be critical to solar cell performance. The effect of band bending on solar cell performance and the dependence of cell performance on {rho} and {phi}{sub TCO} were investigated in detail. Eventually, suggestive design parameters for HIT solar cell fabrication are proposed. (author)

14 SOLAR ENERGY↗

A New Semistructured Algebraic Multigrid Method

Multigrid methods are well suited to large massively parallel computer architectures because they are mathematically optimal and display good parallelization properties. Since current architecture trends are favoring regular compute patterns to achieve high performance, the ability to express structure has become much more important. The hypre software library provides high-performance multigrid preconditioners and solvers through conceptual interfaces, including a semistructured interface that describes matrices primarily in terms of stencils and logically structured grids. This paper presents a new semistructured algebraic multigrid (SSAMG) method built on this interface. The numerical convergence and performance of a CPU implementation of this method are evaluated for a set of semistructured problems. In conclusion, SSAMG achieves significantly better setup times than hypre’s unstructured AMG solvers and comparable convergence. In addition, the new method is capable of solving more complex problems than hypre’s structured solvers.

97 MATHEMATICS AND COMPUTING↗

New Automotive Air Conditioning System Simulation Tool Developed in MATLAB/Simulink

Further improvements in vehicle fuel efficiency require accurate evaluation of the vehicle's transient total power requirement. When operated, the air conditioning (A/C) system is the largest auxiliary load on a vehicle; therefore, accurate evaluation of the load it places on the vehicle's engine and/or energy storage system is especially important. Vehicle simulation software, such as 'Autonomie,' has been used by OEMs to evaluate vehicles' energy performance. A transient A/C simulation tool incorporated into vehicle simulation models would also provide a tool for developing more efficient A/C systems through a thorough consideration of the transient A/C system performance. The dynamic system simulation software Matlab/Simulink was used to develop new and more efficient vehicle energy system controls. The various modeling methods used for the new simulation tool are described in detail. Comparison with measured data is provided to demonstrate the validity of the model.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗

IDAC-Bio, A Software for Internal Dosimetry Based on the New ICRP Biokinetic Models and Specific Absorbed Fractions

Radiation dosimetry is central to virtually all radiation safety applications, optimization, and research. It relates to various individuals and population groups and to miscellaneous exposure situations—including planned, existing, and emergency situations. The International Commission on Radiological Protection (ICRP) has developed a new computational framework for internal dose estimations. Important components are more detailed and improved anatomical models and more realistic biokinetic models than before. The ICRP is currently producing new organ dose and effective dose coefficients for occupational intakes of radionuclides (OIR) and environmental intakes of radionuclides (EIR), which supersede the earlier dose coefficients in Publication 68 and the Publication 72 series, respectively. However, the ICRP only publishes dose coefficients for a single acute intake of a radionuclide and for an integration period of 50 years for intake by adults and to age 70 years for intakes by pre-adults. The new software, IDAC-Bio, performs committed absorbed dose and effective dose calculations for a selectable intake scenario, e.g., for a continuous intake or an intake during x hours per day and y days per week, and for any selected integration time. The software uses the primary data and models of the ICRP biokinetic models and numerically solves the biokinetic model and calculates the absorbed doses to organs and tissues in the ICRP reference human phantoms. The software calculates absorbed dose using the nuclear decay data in ICRP publication 107. IDAC-Bio is a further development and an important addition to the internal dosimetry program IDAC-Dose2.1. Here, the results generated by the software were validated against published ICRP dose coefficients. The potential of the software is illustrated by dose calculations for a nuclear power plant worker who had been exposed to varying levels of 60 Co and who had undergone repeated whole-body measurements, and for a hypothetical member of the public subject to future releases of 148 Gd from neutron spallation in tungsten at the European Spallation Source.

61 RADIATION PROTECTION AND DOSIMETRY↗

HPC Colony II: FAST_OS II: Operating Systems and Runtime Systems at Extreme Scale

HPC Colony II has been a 36-month project focused on providing portable performance for leadership class machines—a task made difficult by the emerging variety of more complex computer architectures. The project attempts to move the burden of portable performance to adaptive system software, thereby allowing domain scientists to concentrate on their field rather than the fine details of a new leadership class machine. To accomplish our goals, we focused on adding intelligence into the system software stack. Our revised components include: new techniques to address OS jitter; new techniques to dynamically address load imbalances; new techniques to map resources according to architectural subtleties and application dynamic behavior; new techniques to dramatically improve the performance of checkpoint-restart; and new techniques to address membership service issues at scale.

97 MATHEMATICS AND COMPUTING↗

SAPHIRE 8 Volume 1 - Overview and Summary

The Systems Analysis Programs for Hands-on Integrated Reliability Evaluations (SAPHIRE) is a software application developed for performing a complete probabilistic risk assessment (PRA) using a personal computer (PC) running the Microsoft Windows operating system. SAPHIRE Version 8 is funded by the U.S. Nuclear Regulatory Commission (NRC) and developed by the Idaho National Laboratory (INL). INL's primary role in this project is that of software developer and tester. However, INL also plays an important role in technology transfer by interfacing and supporting SAPHIRE users, who constitute a wide range of PRA practitioners from the NRC, national laboratories, the private sector, and foreign countries. SAPHIRE can be used to model a complex system’s response to initiating events and quantify associated consequential outcome frequencies. Specifically, for nuclear power plant applications, SAPHIRE 8 can identify important contributors to core damage (Level 1 PRA) and containment failure during a severe accident which leads to releases (Level 2 PRA). It can be used for a PRA where the reactor is at full power, low power, or at shutdown conditions. Furthermore, it can be used to analyze both internal and external initiating events and has special features for managing models such as flooding and fire. It can also be used in a limited manner to quantify risk in terms of release consequences to the public and environment (Level 3 PRA). In SAPHIRE 8, the act of creating a model has been separated from the analysis of that model in order to improve the quality of both the model (e.g., by avoiding inadvertent changes) and the analysis. Consequently, in SAPHIRE 8, the analysis of models is performed by using what are called Workspaces. Currently, there are Workspaces for three types of analyses: (1) the NRC’s Accident Sequence Precursor program, where the workspace is called “Events and Condition Assessment (ECA);” (2) the NRC’s Significance Determination Process (SDP); and (3) the General Analysis (GA) workspace. Workspaces are independent of each other and modifications or calculations made within one workspace will not affect another. In addition, each workspace has a user interface and reports tailored for their intended uses. This report provides an overview of the functions and features available in SAPHIRE 8 and presents general instructions for using the software. Since SAPHIRE 8 expands upon Version 7, new and improved features will be discussed.

99 GENERAL AND MISCELLANEOUS↗

LibraryX: A Framework for Cross-Library-Call Optimization

Scientific applications utilize performance libraries as a software engineering concept: these libraries encapsulate important and well-understood (mathematical) operations, allow for reuse, and are implemented and tuned by experts. Domain scientists then implement complex algorithms based on these domainspecific libraries. While individual library calls are optimized, larger performance gains across sequences of calls—sometimes spanning multiple libraries—are often unrealized, forcing a trade-off between performance and implementation complexity.To overcome this issue, we propose LibraryX, an approach and a system that allows for cross-library-call optimization even when library calls stem from multiple performance libraries. LibraryX annotates library calls with semantic information and optimizes entire directed acyclic graphs (DAGs) of calls dynamically using the SPIRAL code generation system. We demonstrate its effectiveness across a range of memory bound workloads, achieving significant speedups on Nvidia, AMD, and Intel accelerators compared to code using native libraries without cross-call optimization.

Rao, Sanil [Carnegie Mellon University,Department ↗

Reproducibility and Variability of I/O Performance on BG/Q: Lessons Learned from a Data Aggregation Algorithm

Reading and writing data efficiently from different tiers of storage is necessary for most scientific simulations to achieve good performance at scale. Many software solutions have been developed to decrease the I/O bottleneck. One wellknown strategy, in the context of collective I/O operations, is the two-phase I/O scheme. This strategy consists of selecting a subset of processes to aggregate contiguous pieces of data before performing reads/writes. In our previous work, we implemented the two-phase I/O scheme with a MPI-based topology-aware algorithm. Our algorithm showed very good performance at scale compared to the standard I/O libraries such as POSIX I/O and MPI I/O. However, the algorithm had several limitations hindering a satisfying reproducibility of our experiments. In this paper, we extend our work by 1) identifying the obstacles we face to reproduce our experiments and 2) discovering solutions that reduce the unpredictability of our results.

97 MATHEMATICS AND COMPUTING↗

WE-G-BRA-02: SafetyNet: Automating Radiotherapy QA with An Event Driven Framework

Purpose: Quality assurance is an essential task in radiotherapy that often requires many manual tasks. We investigate the use of an event driven framework in conjunction with software agents to automate QA and eliminate wait times. Methods: An in house developed subscription-publication service, EventNet, was added to the Aria OIS to be a message broker for critical events occurring in the OIS and software agents. Software agents operate without user intervention and perform critical QA steps. The results of the QA are documented and the resulting event is generated and passed back to EventNet. Users can subscribe to those events and receive messages based on custom filters designed to send passing or failing results to physicists or dosimetrists. Agents were developed to expedite the following QA tasks: Plan Revision, Plan 2nd Check, SRS Winston-Lutz isocenter, Treatment History Audit, Treatment Machine Configuration. Results: Plan approval in the Aria OIS was used as the event trigger for plan revision QA and Plan 2nd check agents. The agents pulled the plan data, executed the prescribed QA, stored the results and updated EventNet for publication. The Winston Lutz agent reduced QA time from 20 minutes to 4 minutes and provided a more accurate quantitative estimate of radiation isocenter. The Treatment Machine Configuration agent automatically reports any changes to the Treatment machine or HDR unit configuration. The agents are reliable, act immediately, and execute each task identically every time. Conclusion: An event driven framework has inverted the data chase in our radiotherapy QA process. Rather than have dosimetrists and physicists push data to QA software and pull results back into the OIS, the software agents perform these steps immediately upon receiving the sentinel events from EventNet. Mr Keranen is an employee of Varian Medical Systems. Dr. Moran’s institution receives research support for her effort for a linear accelerator QA project from Varian Medical Systems. Other quality projects involving her effort are funded by Blue Cross Blue Shield of Michigan, Breast Cancer Research Foundation, and the NIH.

60 APPLIED LIFE SCIENCES↗