Search NASA⌕ Search

SEARCH · Search NASA

Results for “Computational 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 361 records · Page 20

High Performance Computing Metrics to Enable Application-Platform Communication

Sandia has invested heavily in scientific/engineering application development and in the research, development, and deployment of large scale HPC platforms to support the computational needs of these applications. As application developers continually expand the capabilities of their software and spend more time on performance tuning of applications for these platforms, HPC platform resources are at a premium as they are a heavily shared resource serving the varied needs of many users. To ensure that the HPC platform resources are being used effciently and perform as designed, it is necessary to obtain reliable data on resource utilization that will allow us to investigate the occurrence, severity, and causes of performance-affecting contention between applications. The work presented in this paper was an initial step to determine if resource contention can be understood and minimized through monitoring, modeling, planning and infrastructure. This paper describes the set of metric definitions, identified in this research, that can be used as meaningful and potentially actionable indicators of performance-affecting contention between applications. These metrics were verified using the observed slowdown of IOR, IMB, and CTH in operating scenarios that forced contention. This paper also describes system/application monitoring activities that are critical to distilling vast amounts of data into quantities that hold the key to understanding for an application's performance under production conditions and that will ultimately aid in Sandia's efforts to succeed in extreme-scale computing.

97 MATHEMATICS AND COMPUTING↗

High Performance Computing Facility Operational Assessment, CY 2011 Oak Ridge Leadership Computing Facility

Oak Ridge National Laboratory's Leadership Computing Facility (OLCF) continues to deliver the most powerful resources in the U.S. for open science. At 2.33 petaflops peak performance, the Cray XT Jaguar delivered more than 1.4 billion core hours in calendar year (CY) 2011 to researchers around the world for computational simulations relevant to national and energy security; advancing the frontiers of knowledge in physical sciences and areas of biological, medical, environmental, and computer sciences; and providing world-class research facilities for the nation's science enterprise. Users reported more than 670 publications this year arising from their use of OLCF resources. Of these we report the 300 in this review that are consistent with guidance provided. Scientific achievements by OLCF users cut across all range scales from atomic to molecular to large-scale structures. At the atomic scale, researchers discovered that the anomalously long half-life of Carbon-14 can be explained by calculating, for the first time, the very complex three-body interactions between all the neutrons and protons in the nucleus. At the molecular scale, researchers combined experimental results from LBL's light source and simulations on Jaguar to discover how DNA replication continues past a damaged site so a mutation can be repaired later. Other researchers combined experimental results from ORNL's Spallation Neutron Source and simulations on Jaguar to reveal the molecular structure of ligno-cellulosic material used in bioethanol production. This year, Jaguar has been used to do billion-cell CFD calculations to develop shock wave compression turbo machinery as a means to meet DOE goals for reducing carbon sequestration costs. General Electric used Jaguar to calculate the unsteady flow through turbo machinery to learn what efficiencies the traditional steady flow assumption is hiding from designers. Even a 1% improvement in turbine design can save the nation billions of gallons of fuel.

99 GENERAL AND MISCELLANEOUS↗

High performance computation of residual stress and distortion in laser welded 301L stainless sheets

Transient thermo-mechanical simulation of stainless plate laser welding process was performed by a highly efficient and accurate approach-hybrid iterative substructure and adaptive mesh method. Especially, residual stress prediction was enhanced by considering various heat effects in the numerical model. The influence of laser welding heat input on residual stress and welding distortion of stainless thin sheets were investigated by experiment and simulation. X-ray diffraction (XRD) and contour method were used to measure the surficial and internal residual stress respectively. Effect of strain hardening, annealing and melting on residual stress prediction was clarified through a parametric study. It was shown that these heat effects must be taken into account for accurate prediction of residual stresses in laser welded stainless sheets. Reasonable agreement among residual stresses by numerical method, XRD and contour method was obtained. Buckling type welding distortion was also well reproduced by the developed thermo-mechanical FEM.

36 MATERIALS SCIENCE↗

Novel usage of deep learning and high-performance computing in long-baseline neutrino oscillation experiments

Mención Internacional en el título de doctorDeep-learning methods are playing a crucial role in numerous scientific and industrialapplications. Over the past two decades, these techniques have helped in the collection,reconstruction, and analysis of large data samples in particle physics experiments. Themain topic of this PhD research is the study of deep-learning techniques in long-baselineneutrino oscillation experiments. Neutrinos are mysterious light elementary particles,and their investigation is essential to shed light on some of the remaining open questionsin physics. The work presented here describes an algorithm based on a convolutionalneural network developed to provide highly accurate and efficient selections of electronneutrino and muon neutrino interactions in the Deep Underground Neutrino Experiment(DUNE). With this algorithm, the electron neutrino (antineutrino) selection efficiencypeaks at 90% (94%) and exceeds 85% (90%) for reconstructed neutrino energies between2-5 GeV. The selection efficiency for muon neutrino (antineutrino) interactions is foundto have a maximum of 96% (97%) and exceeds 90% (95%) efficiency for reconstructedneutrino energies above 2 GeV. When considering all electron neutrino and antineutrinointeractions as signal (both those appearing from oscillations and those intrinsic tothe beam), a selection purity of 90% is achieved. These event selections are criticalto maximise the sensitivity of the experiment to CP-violating effects, key to furtherunderstand the matter-antimatter asymmetry of the Universe.In high-energy physics experiments, deep learning has also been explored for producingfast simulations and physically-motivated manipulations of simulated images. Some ofthose simulations, such as the light production and detection, are very computationallyexpensive and require novel methods to produce the necessary samples while controllingthe varied underlying physics model parameters. To do so, we invented the model-assistedgenerative adversarial network (MAGAN), first validated on simple generic case studiesand then successfully applied to the DUNE photon-detector simulation.Moreover, we also developed graph neural networks for 3D-voxel classification ofambiguities and optical crosstalk for a different particle physics experiment, most preciselyfor the proposed SuperFGD. This novel 3D-granular plastic-scintillator neutrino detectorwill be used to upgrade the near detector of the T2K neutrino oscillation experiment, and our method reports efficiencies and purities of 94-96% per event in the classificationof particle track voxels.Due to the growth and complexity of deep neural networks, researchers have beeninvestigating techniques to train those networks in a more computationally-efficient way.Many efforts have been made by the community to optimise deep-learning models byparallelising or distributing their training computation across multiple devices. In thisthesis, we study an approach based on data locality for those neural networks that cannotbenefit from scaling their computation due to a significant bottleneck in the data I/O.The research also includes a detailed study on the performance of deep neural networkson hardware accelerator boards.Los métodos de aprendizaje profundo son cada vez más utilizados en numerosas aplicacionescientíficas e industriales hoy en día. Durante las dos últimas décadas, estastécnicas se han empleado en la recolección, reconstrucción y análisis de la gran cantidadde datos generados por experimentos de física de partículas. El tema principal de estatesis doctoral es el uso de estos modelos de aprendizaje profundo en experimentos defísica de neutrinos, en concreto en los experimentos de larga distancia DUNE y T2K. Losneutrinos, partículas fundamentales neutras, de las más ligeras del Universo, pueden serclave para explicar algunas de las cuestiones todavía sin resolver en física fundamental.Entre las diferentes contribuciones que esta tesis ha hecho a su estudio, cabe destacar eldesarrollo de un algoritmo basado en una red de neuronas convolucional para seleccionarcon gran eficiencia y precisión las interacciones de neutrinos electrónicos y muónicos enel Deep Underground Neutrino Experiment (DUNE). La eficiencia de selección obtenidapara neutrinos (antineutrinos) electrónicos alcanza un máximo del 90% (94%) y supera el85% (90%) para neutrinos con energías reconstruidas en el rango 2-5 GeV. La selección deneutrinos (antineutrinos) muónicos tiene una eficiencia máxima del 96% (97%) y excedeel 90% (95%) para neutrinos con energías reconstruidas de más de 2 GeV. Considerandocomo señal todas las interacciones de neutrinos y antineutrinos electrónicos (procedentestanto de oscilaciones como intrínsecos en el haz inicial), se logra una pureza en la seleccióndel 90%. Dichas selecciones de eventos son fundamentales para maximizar la sensibilidaddel experimento a los efectos de violació...

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

Classifier-guided sampling for discrete variable, discontinuous design space exploration: Convergence and computational performance

A classifier-guided sampling (CGS) method is introduced for solving engineering design optimization problems with discrete and/or continuous variables and continuous and/or discontinuous responses. The method merges concepts from metamodel-guided sampling and population-based optimization algorithms. The CGS method uses a Bayesian network classifier for predicting the performance of new designs based on a set of known observations or training points. Unlike most metamodeling techniques, however, the classifier assigns a categorical class label to a new design, rather than predicting the resulting response in continuous space, and thereby accommodates nondifferentiable and discontinuous functions of discrete or categorical variables. The CGS method uses these classifiers to guide a population-based sampling process towards combinations of discrete and/or continuous variable values with a high probability of yielding preferred performance. Accordingly, the CGS method is appropriate for discrete/discontinuous design problems that are ill-suited for conventional metamodeling techniques and too computationally expensive to be solved by population-based algorithms alone. In addition, the rates of convergence and computational properties of the CGS method are investigated when applied to a set of discrete variable optimization problems. Results show that the CGS method significantly improves the rate of convergence towards known global optima, on average, when compared to genetic algorithms.

42 ENGINEERING↗

Algorithms for high-throughput disk-to-disk sorting, In: SC '13 Proceedings of the International Conference on High Performance Computing, Networking, Storage and Analysis

In this paper, we present a new out-of-core sort algorithm, designed for problems that are too large to fit into the aggregate RAM available on modern supercomputers. We analyze the performance including the cost of IO and demonstrate the fastest (to the best of our knowledge) reported throughput using the canonical sortBenchmark on a general-purpose, production HPC resource running Lustre. By clever use of available storage and a formulation of asynchronous data transfer mechanisms, we are able to almost completely hide the computation (sorting) behind the IO latency. This latency hiding enables us to achieve comparable execution times, including the additional temporary IO required, between a large sort problem (5TB) run as a single, in-RAM sort and our out-of-core approach using 1/10th the amount of RAM. In our largest run, sorting 100TB of records using 1792 hosts, we achieved an end-to-end throughput of 1.24TB/min using our general-purpose sorter, improving on the current Daytona record holder by 65%.

Computer Science↗

Application Characterization at Scale: Lessons learned from developing a distributed Open Community Runtime system for High Performance Computing

Since 2012, the U.S. Department of Energy’s X-Stack program has been developing solutions including runtime systems, programming models, languages, compilers, and tools for the Exascale system software to address crucial performance and power requirements. Fine grain programming models and runtime systems show a great potential to efficiently utilize the underlying hardware. Thus, they are essential to many X-Stack efforts. An abundant amount of small tasks can better utilize the vast parallelism available on current and future machines. Moreover, finer tasks can recover faster and adapt better, due to a decrease in state and control. Nevertheless, current applications have been written to exploit old paradigms (such as Communicating Sequential Processor and Bulk Synchronous Parallel processing). To fully utilize the advantages of these new systems, applications need to be adapted to these new paradigms. As part of the applications’ porting process, in-depth characterization studies, focused on both application characteristics and runtime features, need to take place to fully understand the application performance bottlenecks and how to resolve them. This paper presents a characterization study for a novel high performance runtime system, called the Open Community Runtime, using key HPC kernels as its vehicle. This study has the following contributions: one of the first high performance, fine grain, distributed memory runtime system implementing the OCR standard (version 0.99a); and a characterization study of key HPC kernels in terms of runtime primitives running on both intra and inter node environments. Running on a general purpose cluster, we have found up to 1635x relative speed-up for a parallel tiled Cholesky Kernels on 128 nodes with 16 cores each and a 1864x relative speed-up for a parallel tiled Smith-Waterman kernel on 128 nodes with 30 cores.

Application characterization↗

HACC: extreme scaling and performance across diverse architectures, In: SC '13 Proceedings of the International Conference on High Performance Computing, Networking, Storage and Analysis

Supercomputing is evolving towards hybrid and accelerator-based architectures with millions of cores. The HACC (Hardware/Hybrid Accelerated Cosmology Code) framework exploits this diverse landscape at the largest scales of problem size, obtaining high scalability and sustained performance. Developed to satisfy the science requirements of cosmological surveys, HACC melds particle and grid methods using a novel algorithmic structure that flexibly maps across architectures, including CPU/GPU, multi/many-core, and Blue Gene systems. We demonstrate the success of HACC on two very different machines, the CPU/GPU system Titan and the BG/Q systems Sequoia and Mira, attaining unprecedented levels of scalable performance. We demonstrate strong and weak scaling on Titan, obtaining up to 99.2% parallel efficiency, evolving 1.1 trillion particles. On Sequoia, we reach 13.94 PFlops (69.2% of peak) and 90% parallel efficiency on 1,572,864 cores, with 3.6 trillion particles, the largest cosmological benchmark yet performed. HACC design concepts are applicable to several other supercomputer applications.

Computer Science↗

Integration of Russian Tier-1 Grid Center with High Performance Computers at NRC-KI for LHC experiments and beyond HENP

The LHC experiments are preparing for the precision measurements and further discoveries that will be made possible by higher LHC energies from April 2015 (LHC Run2). The need for simulation, data processing and analysis would overwhelm the expected capacity of grid infrastructure computing facilities deployed by the Worldwide LHC Computing Grid (WLCG). To meet this challenge the integration of the opportunistic resources into LHC computing model is highly important. The Tier-1 facility at Kurchatov Institute (NRC-KI) in Moscow is a part of WLCG and it will process, simulate and store up to 10% of total data obtained from ALICE, ATLAS and LHCb experiments. In addition Kurchatov Institute has supercomputers with peak performance 0.12 PFLOPS. The delegation of even a fraction of supercomputing resources to the LHC Computing will notably increase total capacity. In 2014 the development a portal combining a Tier-1 and a supercomputer in Kurchatov Institute was started to provide common interfaces and storage. The portal will be used not only for HENP experiments, but also by other data- and compute-intensive sciences like biology with genome sequencing analysis; astrophysics with cosmic rays analysis, antimatter and dark matter search, etc.

97 MATHEMATICS AND COMPUTING↗

Modeling and Predicting Performance of High Performance Computing Applications on Hardware Accelerators

Computers with hardware accelerators, also referred to as hybrid-core systems, speedup applications by offloading certain compute operations that can run faster on accelerators. Thus, it is not surprising that many of top500 supercomputers use accelerators. However, in addition to procurement cost, significant programming and porting effort is required to realize the potential benefit of such accelerators. Hence, before building such a system it is prudent to answer the question 'what is the projected performance benefit from accelerators for the workloads of interest?'. We address this question by way of a performance-modeling framework that predicts realizable application performance on accelerators rapidly and accurately without going to the considerable effort of porting and tuning. The modeling framework first automatically identifies commonly found compute patterns in scientific applications which we term idioms, which may benefit by accelerator technology. Next the framework models the predicted speedup of those idioms if they were to be ported to and run on hardware accelerators. As a proof of concept we characterize two kinds of accelerators 1) the FPGA accelerators on a Convey HC-1 system and 2) an NVIDIA FERMI GPU accelerator. We model performance of the idioms gather/scatter and stream and our predictions show that where these occur in two full-scale HPC applications, Milc and HYCOM, gather/scatter speeds up by as much as 15X, and stream by as much as 14X, whereas the overall compute time of Milc improves by 3.4% and HYCOM by 20%.

Computer Science↗

Newmark local time stepping on high-performance computing architectures

In multi-scale complex media, finite element meshes often require areas of local refinement, creating small elements that can dramatically reduce the global time-step for wave-propagation problems due to the CFL condition. Local time stepping (LTS) algorithms allow an explicit time-stepping scheme to adapt the time-step to the element size, allowing near-optimal time-steps everywhere in the mesh. We develop an efficient multilevel LTS-Newmark scheme and implement it in a widely used continuous finite element seismic wave-propagation package. In particular, we extend the standard LTS formulation with adaptations to continuous finite element methods that can be implemented very efficiently with very strong element-size contrasts (more than 100x). Capable of running on large CPU and GPU clusters, we present both synthetic validation examples and large scale, realistic application examples to demonstrate the performance and applicability of the method and implementation on thousands of CPU cores and hundreds of GPUs.

97 MATHEMATICS AND COMPUTING↗

Checkpointing Strategies for Shared High-Performance Computing Platforms

Input/output (I/O) from various sources often contend for scarcely available bandwidth. For example, checkpoint/restart (CR) protocols can help to ensure application progress in failure-prone environments. However, CR I/O alongside an application's normal, requisite I/O can increase I/O contention and might negatively impact performance. In this work, we consider different aspects (system-level scheduling policies and hardware) that optimize the overall performance of concurrently executing CR-based applications that share I/O resources. We provide a theoretical model and derive a set of necessary constraints to minimize the global waste on a given platform. Our results demonstrate that Young/Daly's optimal checkpoint interval, despite providing a sensible metric for a single, undisturbed application, is not sufficient to optimally address resource contention at scale. We show that by combining optimal checkpointing periods with contention-aware system-level I/O scheduling strategies, we can significantly improve overall application performance and maximize the platform throughput. Finally, we evaluate how specialized hardware, namely burst buffers, may help to mitigate the I/O contention problem. Altogether, these results provide critical analysis and direct guidance on how to design efficient, CR ready, large -scale platforms without a large investment in the I/O subsystem.

97 MATHEMATICS AND COMPUTING↗

Predictive Dynamic Simulation for Large-Scale Power Systems through High-Performance Computing

Power system dynamic simulation solves a set of differential-algebraic equations to determine the time-series trajectory when the system is subject to disturbances such as a short-circuit fault, generator tripping, or line switching. Due to computational inefficiency, dynamic simulation, though widely used for off-line studies, has not been used in real-time operation. That limits the ability to operate a much-evolved power system with significant dynamic and stochastic behaviors introduced by the increasing penetration of renewable generation and the deployment of smart grid technologies. The need for performing dynamic simulation in real-time or faster than real-time for power grid operation becomes apparent. And such predictive dynamic simulation can enable many new power grid operation functions such as real-time path rating. To improve the computational efficiency of dynamic simulation requires parallel computing implementation of the solution methods, as computers no longer have a single core. This paper examines the equations of power system dynamic simulation and implements a parallel version of dynamic simulation. The testing results clearly show the performance has been significantly improved, and dynamic simulation of a large-scale power system with a size equivalent to the Western U.S. power grid achieves, for the first time, a performance of three times faster than real time. The potential of predictive dynamic simulation is also discussed.

differential algebraic equations↗

Teuchos C++ memory management classes, idioms, and related topics, the complete reference : a comprehensive strategy for safe and efficient memory management in C++ for high performance computing.

The ubiquitous use of raw pointers in higher-level code is the primary cause of all memory usage problems and memory leaks in C++ programs. This paper describes what might be considered a radical approach to the problem which is to encapsulate the use of all raw pointers and all raw calls to new and delete in higher-level C++ code. Instead, a set of cooperating template classes developed in the Trilinos package Teuchos are used to encapsulate every use of raw C++ pointers in every use case where it appears in high-level code. Included in the set of memory management classes is the typical reference-counted smart pointer class similar to boost::shared ptr (and therefore C++0x std::shared ptr). However, what is missing in boost and the new standard library are non-reference counted classes for remaining use cases where raw C++ pointers would need to be used. These classes have a debug build mode where nearly all programmer errors are caught and gracefully reported at runtime. The default optimized build mode strips all runtime checks and allows the code to perform as efficiently as raw C++ pointers with reasonable usage. Also included is a novel approach for dealing with the circular references problem that imparts little extra overhead and is almost completely invisible to most of the code (unlike the boost and therefore C++0x approach). Rather than being a radical approach, encapsulating all raw C++ pointers is simply the logical progression of a trend in the C++ development and standards community that started with std::auto ptr and is continued (but not finished) with std::shared ptr in C++0x. Using the Teuchos reference-counted memory management classes allows one to remove unnecessary constraints in the use of objects by removing arbitrary lifetime ordering constraints which are a type of unnecessary coupling [23]. The code one writes with these classes will be more likely to be correct on first writing, will be less likely to contain silent (but deadly) memory usage errors, and will be much more robust to later refactoring and maintenance. The level of debug-mode runtime checking provided by the Teuchos memory management classes is stronger in many respects than what is provided by memory checking tools like Valgrind and Purify while being much less expensive. However, tools like Valgrind and Purify perform a number of types of checks (like usage of uninitialized memory) that makes these tools very valuable and therefore complement the Teuchos memory management debug-mode runtime checking. The Teuchos memory management classes and idioms largely address the technical issues in resolving the fragile built-in C++ memory management model (with the exception of circular references which has no easy solution but can be managed as discussed). All that remains is to teach these classes and idioms and expand their usage in C++ codes. The long-term viability of C++ as a usable and productive language depends on it. Otherwise, if C++ is no safer than C, then is the greater complexity of C++ worth what one gets as extra features? Given that C is smaller and easier to learn than C++ and since most programmers don't know object-orientation (or templates or X, Y, and Z features of C++) all that well anyway, then what really are most programmers getting extra out of C++ that would outweigh the extra complexity of C++ over C? C++ zealots will argue this point but the reality is that C++ popularity has peaked and is becoming less popular while the popularity of C has remained fairly stable over the last decade22. Idioms like are advocated in this paper can help to avert this trend but it will require wide community buy-in and a change in the way C++ is taught in order to have the greatest impact. To make these programs more secure, compiler vendors or static analysis tools (e.g. klocwork23) could implement a preprocessor-like language similar to OpenMP24 that would allow the programmer to declare (in comments) that certain blocks of code should be ''pointer-free'' or allow smaller blocks to be 'pointers allowed'. This would significantly improve the robustness of code that uses the memory management classes described here.

99 GENERAL AND MISCELLANEOUS↗

High Performance Computing Facility Operational Assessment, FY 2010 Oak Ridge Leadership Computing Facility

Oak Ridge National Laboratory's (ORNL's) Cray XT5 supercomputer, Jaguar, kicked off the era of petascale scientific computing in 2008 with applications that sustained more than a thousand trillion floating point calculations per second - or 1 petaflop. Jaguar continues to grow even more powerful as it helps researchers broaden the boundaries of knowledge in virtually every domain of computational science, including weather and climate, nuclear energy, geosciences, combustion, bioenergy, fusion, and materials science. Their insights promise to broaden our knowledge in areas that are vitally important to the Department of Energy (DOE) and the nation as a whole, particularly energy assurance and climate change. The science of the 21st century, however, will demand further revolutions in computing, supercomputers capable of a million trillion calculations a second - 1 exaflop - and beyond. These systems will allow investigators to continue attacking global challenges through modeling and simulation and to unravel longstanding scientific questions. Creating such systems will also require new approaches to daunting challenges. High-performance systems of the future will need to be codesigned for scientific and engineering applications with best-in-class communications networks and data-management infrastructures and teams of skilled researchers able to take full advantage of these new resources. The Oak Ridge Leadership Computing Facility (OLCF) provides the nation's most powerful open resource for capability computing, with a sustainable path that will maintain and extend national leadership for DOE's Office of Science (SC). The OLCF has engaged a world-class team to support petascale science and to take a dramatic step forward, fielding new capabilities for high-end science. This report highlights the successful delivery and operation of a petascale system and shows how the OLCF fosters application development teams, developing cutting-edge tools and resources for next-generation systems.

29 ENERGY PLANNING, POLICY, AND ECONOMY↗

The Kokkos EcoSystem: Comprehensive Performance Portability For High Performance Computing

State of the art Engineering and Science codes have grown in complexity dramatically over the last two decades. As a consequence application teams have adopted more sophisticated development strategies, leveraging third party libraries, deploying comprehensive testing and using advanced debugging and profiling tools. In todays environment of diverse hardware platforms, these applications also desire performance portability - avoiding the need to duplicate work for various platforms - which makes it necessary that these tools and libraries also work across the various systems. The Kokkos EcoSystem provides that portable software stack. Based on the Kokkos Core Programming Model, the EcoSystem provides math libraries, interoperability capabilities with Python and Fortran, and Tools for analysing, debugging, and optimizing applications. In this paper we will provide an overview of the components, discuss some specific use cases, and highlight how co-designing these components enables a more developer friendly experience.

42 ENGINEERING↗