Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “EXASCALE COMPUTING”

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 559 records · Page 31

The Ristra Project: FY20/21 Milestone Report

The ASC Advanced Technology Development and Mitigation (ATDM) sub-program was established in 2014 to develop new simulation tools operating on exascale-class computers to serve NNSA (see Appendix B). Over the course of ATDM, LANL management have set a strategy for exascale-class application codes that follows two supportive and mutually risk-mitigating paths: evolution for established production integrated design codes (IDCs) – with a strong pedigree within the user community – based upon existing programming paradigms(MPI+X); and a new start ATDM project, Ristra, a high-risk/high-reward push for a next-generation multi-physics, multi-scale simulation toolkit based on emerging advanced programming systems(with an initial focus on data-flow task-based models exemplified by Legion). The role of Ristra as the high-risk/high-reward path for LANL’s codes was fully consistent with the goals of ATDM as described in Appendix B, in particular its emphasis on evolving ASC capabilities through novel computing programming models and computing technologies.

97 MATHEMATICS AND COMPUTING↗

Matrix Multiply Performance of GPUs on Exascale-class HPE/Cray Systems

The computation of dense matrix-matrix products (GEMMs) is central to many modeling and simulation workloads as well as AI/ML deep learning campaigns. In fact, millions of dollars are spent annually on computing GEMMs, and large model training demands are increasing exponentially. Specialized processors such as GPUs are designed to perform well for these operations. However, the performance of GEMMs on GPUs can exhibit complex behaviors depending on many factors, making it challenging to optimize the performance of GEMMs on these processors. In this study we undertake an examination of GEMM performance on several leading GPU models taken from product lines of GPUs to be deployed in forthcoming exascale computing systems. We show results to illustrate the many factors that can affect performance of GEMMs on GPUs. We then present data collected from a large number of test runs for an example GEMM operation to show the dependence behaviors of GEMM rate on matrix dimensions. Finally, we show results from machine learning-based performance models using novel feature engineering methods to fit the measured performance, providing a potential basis for GEMM performance tuning and autotuning methods for GPUs. Recommendations are also given for how to achieve high GEMM performance on modern GPUs.

Melesse Vergara, Veronica↗

Multiphysics Time-Integration for Turbulent Combustion at the Exascale

Turbulent reacting flow systems are often modeled with coupled time-dependent partial differential equations (PDEs). Solving such equations can easily tax the world's largest supercomputers. One pragmatic strategy for attacking such problems is to split the PDEs into components that can more easily be solved in isolation. This generic operator-splitting strategy leads to a set of ordinary differential equations (ODEs) that need to be solved as part of an "outer-loop" time-stepping approach. In many combustion applications, the ODEs to be solved can be very stiff, exhibiting timescales that span many orders of magnitude. The SUNDIALS library provides a plethora of robust time integration algorithms for solving these ODEs on exascale-capable computing hardware, yet for many complex applications (such multicomponent fuels or emissions predictions), the chemical models remain too complex to solve using reasonable resources. The Quasi-Steady State Approximation (QSSA) can be an effective tool for reducing the size and stiffness of the simulations. In this talk, I will discuss the use of the SUDIALS library of ODE solvers together with automatic code generation tools to solve complex turbulent reacting flow problems using QSSA models.

chemistry↗

PluginPlay: Enabling exascale scientific software one module at a time

For many computational chemistry packages, being able to efficiently and effectively scale across an exascale cluster is a heroic feat. Collective experience from the Department of Energy’s Exascale Computing Project suggests that achieving exascale performance requires far more planning, design, and optimization than scaling to petascale. In many cases, entire rewrites of software are necessary to address fundamental algorithmic bottlenecks. This in turn requires a tremendous amount of resources and development time, resources that cannot reasonably be afforded by every computational science project. It thus becomes imperative that computational science transition to a more sustainable paradigm. Key to such a paradigm is modular software. While the importance of modular software is widely recognized, what is perhaps not so widely appreciated is the effort still required to leverage modular software in a sustainable manner. The present manuscript introduces PluginPlay, https://github.com/NWChemEx-Project/PluginPlay, an inversion-of-control framework designed to facilitate developing, maintaining, and sustaining modular scientific software packages. Here this manuscript focuses on the design aspects of PluginPlay and how they specifically influence the performance of the resulting package. Although, PluginPlay serves as the framework for the NWChemEx package, PluginPlay is not tied to NWChemEx or even computational chemistry. We thus anticipate PluginPlay to prove to be a generally useful tool for a number of computational science packages looking to transition to the exascale.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Crosscut report: Exascale Requirements Reviews, March 9–10, 2017 – Tysons Corner, Virginia. An Office of Science review sponsored by: Advanced Scientific Computing Research, Basic Energy Sciences, Biological and Environmental Research, Fusion Energy Sciences, High Energy Physics, Nuclear Physics

The mission of the U.S. Department of Energy Office of Science (DOE SC) is the delivery of scientific discoveries and major scientific tools to transform our understanding of nature and to advance the energy, economic, and national security missions of the United States. To achieve these goals in today’s world requires investments in not only the traditional scientific endeavors of theory and experiment, but also in computational science and the facilities that support large-scale simulation and data analysis. The Advanced Scientific Computing Research (ASCR) program addresses these challenges in the Office of Science. ASCR’s mission is to discover, develop, and deploy computational and networking capabilities to analyze, model, simulate, and predict complex phenomena important to DOE. ASCR supports research in computational science, three high-performance computing (HPC) facilities — the National Energy Research Scientific Computing Center (NERSC) at Lawrence Berkeley National Laboratory and Leadership Computing Facilities at Argonne (ALCF) and Oak Ridge (OLCF) National Laboratories — and the Energy Sciences Network (ESnet) at Berkeley Lab. ASCR is guided by science needs as it develops research programs, computers, and networks at the leading edge of technologies. As we approach the era of exascale computing, technology changes are creating challenges for science programs in SC for those who need to use high performance computing and data systems effectively. Numerous significant modifications to today’s tools and techniques will be needed to realize the full potential of emerging computing systems and other novel computing architectures. To assess these needs and challenges, ASCR held a series of Exascale Requirements Reviews in 2015–2017, one with each of the six SC program offices,1 and a subsequent Crosscut Review that sought to integrate the findings from each. Participants at the reviews were drawn from the communities of leading domain scientists, experts in computer science and applied mathematics, ASCR facility staff, and DOE program managers in ASCR and the respective program offices. The purpose of these reviews was to identify mission-critical scientific problems within the DOE Office of Science (including experimental facilities) and determine the requirements for the exascale ecosystem that would be needed to address those challenges. The exascale ecosystem includes exascale computing systems, high-end data capabilities, efficient software at scale, libraries, tools, and other capabilities. This effort will contribute to the development of a strategic roadmap for ASCR compute and data facility investments and will help the ASCR Facility Division establish partnerships with Office of Science stakeholders. It will also inform the Office of Science research needs and agenda. The results of the six reviews have been published in reports available on the web at http://exascaleage.org/. This report presents a summary of the individual reports and of common and crosscutting findings, and it identifies opportunities for productive collaborations among the DOE SC program offices.

97 MATHEMATICS AND COMPUTING↗

High-Order/Low-Order methods for ocean modeling

We examine a High Order/Low Order (HOLO) approach for a z-level ocean model and show that the traditional semi-implicit and split-explicit methods, as well as a recent preconditioning strategy, can easily be cast in the framework of HOLO methods. The HOLO formulation admits an implicit-explicit method that is algorithmically scalable and second-order accurate, allowing timesteps much larger than the barotropic time scale. We demonstrate how HOLO approaches, in particular the implicit-explicit method, can provide a solid route for ocean simulation to heterogeneous computing and exascale environments.

16 TIDAL AND WAVE POWER↗

ExaWind: Exascale Predictive Wind Plant Flow Physics Modeling

The scientific goal of the ExaWind project is to advance our fundamental understanding of the flow physics governing whole wind plant performance, including wake formation, complex terrain impacts, and turbine-turbine-interaction effects. The primary application codes in the ExaWind environment are Nalu-Wind, an unstructured-grid computational fluid dynamics (CFD) code, AMR-Wind, a structured-grid CFD code, and OpenFAST, a whole-turbine simulation code. In this poster we present our progress towards simulating the ExaWind challenge problem, which is a predictive simulation of a wind farm with tens of megawatt-scale wind turbines dispersed over an area of 50 square kilometers.

49 EE - Wind and Water Power Program - Wind (EE-4W↗

Integrating ytopt and libEnsemble to autotune OpenMC

Ytopt is a Python machine-learning-based autotuning software package developed within the ECP PROTEAS-TUNE project. The ytopt software adopts an asynchronous search framework that consists of sampling a small number of input parameter configurations and progressively fitting a surrogate model over the input-output space until exhausting the user-defined maximum number of evaluations or the wall-clock time. libEnsemble is a Python toolkit for coordinating workflows of asynchronous and dynamic ensembles of calculations across massively parallel resources developed within the ECP PETSc/TAO project. libEnsemble helps users take advantage of massively parallel resources to solve design, decision, and inference problems and expands the class of problems that can benefit from increased parallelism. In this paper we present our methodology and framework to integrate ytopt and libEnsemble to take advantage of massively parallel resources to accelerate the autotuning process. Specifically, we focus on using the proposed framework to autotune the ECP ExaSMR application OpenMC, an open source Monte Carlo particle transport code. OpenMC has seven tunable parameters some of which have large ranges such as the number of particles in-flight, which is in the range of 100,000 to 8 million, with its default setting of 1 million. Setting the proper combination of these parameter values to achieve the best performance is extremely time-consuming. Therefore, we apply the proposed framework to autotune the MPI/OpenMP offload version of OpenMC based on a user-defined metric such as the figure of merit (FoM) (particles/s) or energy efficiency energy-delay product (EDP) on Crusher at Oak Ridge Leadership Computing Facility. In conclusion, the experimental results show that we achieve the improvement up to 29.49% in FoM and up to 30.44% in EDP.

Autotuning↗

Scalable Stochastic Transmission Expansion: A Use Case for ExaSGD

The intermittent nature of renewable energy poses new challenges for power grids due to its variable and un- certain power output. These features of renewable generation are becoming more relevant to transmission planning as grids reach higher penetration levels of renewable energy. In this paper we present an approach for transmission planning based on scalable computational approaches which enable the explicit consideration of operational uncertainties in the planning process. Using three-stage stochastic programming and the progressive hedging algorithm, we compute transmission expansion decisions on a modified RTS-GMLC test system. We augment the grid with large amounts of wind generation and consider many operational scenarios subject to wind uncertainty. This is an example of a possible use of the ExaSGD security constrained AC optimal power flow solver.

Exascale Computing Project↗

Application health monitoring for extreme-scale resiliency using cooperative fault management

Resiliency is and will be a critical factor in determining scientific productivity on current and exascale supercomputers, and beyond. Applications oblivious to and incapable of handling transient soft and hard errors could waste supercomputing resources or, worse, yield misleading scientific insights. In this work, we introduce a novel application-driven silent error detection and recovery strategy based on application health monitoring. Our methodology uses application output that follows known patterns, as indicators of an application's health and knowledge that violation of these patterns could be indication of faults. Information from system monitors that report hardware and software health status is used to corroborate faults. Collectively, this information is used by a fault coordinator agent to take preventive and corrective measures by applying computational steering to an application between checkpoints. This cooperative fault management system uses the Fault Tolerance Backplane as a communication channel. The benefits of this framework are demonstrated with two real application case studies, molecular dynamics, and quantum chemistry simulations, on scalable clusters with simulated memory and I/O corruptions. Lastly, the developed approach is general and can be easily applied to other applications.

97 MATHEMATICS AND COMPUTING↗

Scientific Challenges for Understanding the Quantum Universe

A workshop titled "Scientific Challenges for Understanding the Quantum Universe" was held December 9-11, 2008, at the Kavli Institute for Particle Astrophysics and Cosmology at the Stanford Linear Accelerator Center-National Accelerator Laboratory. The primary purpose of the meeting was to examine how computing at the extreme scale can contribute to meeting forefront scientific challenges in particle physics, particle astrophysics and cosmology. The workshop was organized around five research areas with associated panels. Three of these, "High Energy Theoretical Physics," "Accelerator Simulation," and "Experimental Particle Physics," addressed research of the Office of High Energy Physics’ Energy and Intensity Frontiers, while the"Cosmology and Astrophysics Simulation" and "Astrophysics Data Handling, Archiving, and Mining" panels were associated with the Cosmic Frontier.

43 PARTICLE ACCELERATORS↗

Large-Scale Materials Modeling at Quantum Accuracy: Ab Initio Simulations of Quasicrystals and Interacting Extended Defects in Metallic Alloys

Ab initio electronic-structure has remained dichotomous between achievable accuracy and length-scale. Quantum many-body (QMB) methods realize quantum accuracy but fail to scale. Density functional theory (DFT) scales favorably but remains far from quantum accuracy. We present a framework that breaks this dichotomy by use of three interconnected modules: (i) invDFT: a methodological advance in inverse DFT linking QMB methods to DFT; (ii) MLXC: a machine-learned density functional trained with invDFT data, commensurate with quantum accuracy; (iii) DFT-FE-MLXC: an adaptive higher-order spectral finite-element (FE) based DFT implementation that integrates MLXC with efficient solver strategies and HPC innovations in FE-specific dense linear algebra, mixed-precision algorithms, and asynchronous compute-communication. Furthermore, we demonstrate a paradigm shift in DFT that not only provides an accuracy commensurate with QMB methods in ground-state energies, but also attains an unprecedented performance of 659.7 PFLOPS (43.1% peak FP64 performance) on 619,124 electrons using 8,000 GPU nodes of Frontier supercomputer.

density functional theory↗

Basic Research Needs for Microelectronics

Since the invention of the integrated circuit in 1958, advances in microelectronics have followed Moore’s law and other scaling laws. These have led to over a billion-fold increase in the number of transistors on a chip. The Department of Energy’s Office of Science (DOE-SC) programs have always been at the cutting edge of microelectronics, making major contributions to the scientific understanding, materials, and advanced instrumentation that enabled innovations to promote scaling. This has driven transformative advances in microelectronics for the challenging demands of DOE’s high performance computing and science facilities. Now, strong evidence exists that scaling is approaching its physical and economic limits, and yet, the growth of data-centric computing and sensor networks is redefining computing workloads and microelectronics needs. In addition, greatly improved microelectronics are needed for the nation’s electricity grid if it is to be energy-efficient, resilient to natural phenomena and intentional attack, and agile in adapting to fluctuations in demand and power generation. Sustained and rapid progress in microelectronics science and technology from millivolt to megavolt scales is thus essential if we are to continue pushing the boundaries of science within DOE, and more significantly, to continue to lead the global information and power technology revolution. In October 2018, DOE-SC convened a workshop to address Basic Research Needs for Microelectronics by a “co-design” approach. Participants focused on scientific issues associated with advanced microelectronics technologies for applications relevant to the DOE mission, including computing, power grid management, and science facility workloads. (Topics of direct relevance to quantum information science and quantum computing were outside the scope of this workshop.) The workshop was organized around microelectronics needs for three areas: 1) future DOE-SC facilities, 2) high performance computing beyond exascale (a billion billion calculations per second), and 3) power control, conversion, and detection for a modernized electrical grid and related high power applications. Panels were formed around each of these areas, with a fourth panel formed to address Crosscutting Research. Workshop participants were asked to focus on a co-design innovation ecosystem in which materials, chemistries, devices, systems, architectures, algorithms, and software are researched and developed in a closely integrated fashion, with feedback and interdisciplinary collaboration at each interface in this microelectronics technology hierarchy. The panels identified the five Priority Research Directions on the following page.

36 MATERIALS SCIENCE↗

Apparatus and method for fusion of compute and switching functions of exascale system into a single component by using configurable network-on-chip fabric with distributed dual mode input-output ports and programmable network interfaces

Described is an apparatus which comprises: a Network-On-Chip fabric using crossbar switches, having distributed ingress and egress ports; and a dual-mode network interface coupled to at least one crossbar switch, the dual-mode network interface is to include: a dual-mode circuitry; a controller operable to: configure the dual-mode circuitry to transmit and receive differential signals via the egress and ingress ports, respectively, and configure the dual-mode circuitry to transmit and receive signal-ended signals via the egress and ingress ports, respectively.

97 MATHEMATICS AND COMPUTING↗

Online data analysis and reduction: An important co-design motif for extreme-scale computers

A growing disparity between supercomputer computation speeds and I/O rates means that it is rapidly becoming infeasible to analyze supercomputer application output only after that output has been written to a file system. Instead, data-generating applications must run concurrently with data reduction and/or analysis operations, with which they exchange information via high-speed methods such as interprocess communications. The resulting parallel computing motif, online data analysis and reduction (ODAR), has important implications for both application and HPC systems design. Here we introduce the ODAR motif and its co-design concerns, describe a co-design process for identifying and addressing those concerns, present tools that assist in the co-design process, and present case studies to illustrate the use of the process and tools in practical settings.

Data Analysis↗

Scientific Data Services -- A High-Performance I/O System with Array Semantics

As high-performance computing approaches exascale, the existing I/O system design is having trouble keeping pace in both performance and scalability. We propose to address this challenge by adopting database principles and techniques in parallel I/O systems. First, we propose to adopt an array data model because many scientific applications represent their data in arrays. This strategy follows a cardinal principle from database research, which separates the logical view from the physical layout of data. This high-level data model gives the underlying implementation more freedom to optimize the physical layout and to choose the most effective way of accessing the data. For example, knowing that a set of write operations is working on a single multi-dimensional array makes it possible to keep the subarrays in a log structure during the write operations and reassemble them later into another physical layout as resources permit. While maintaining the high-level view, the storage system could compress the user data to reduce the physical storage requirement, collocate data records that are frequently used together, or replicate data to increase availability and fault-tolerance. Additionally, the system could generate secondary data structures such as database indexes and summary statistics. We expect the proposed Scientific Data Services approach to create a “live” storage system that dynamically adjusts to user demands and evolves with the massively parallel storage hardware.

97 MATHEMATICS AND COMPUTING↗

Investigating power capping toward energy-efficient scientific applications: Investigating Power Capping toward Energy-Efficient Scientific Applications

The emergence of power efficiency as a primary constraint in processor and system design poses new challenges concerning power and energy awareness for numerical libraries and scientific applications. Power consumption also plays a major role in the design of data centers, which may house petascale or exascale-level computing systems. At these extreme scales, understanding and improving the energy efficiency of numerical libraries and their related applications becomes a crucial part of the successful implementation and operation of the computing system. In this paper, we study and investigate the practice of controlling a compute system's power usage, and we explore how different power caps affect the performance of numerical algorithms with different computational intensities. Further, we determine the impact, in terms of performance and energy usage, that these caps have on a system running scientific applications. This analysis will enable us to characterize the types of algorithms that benefit most from these power management schemes. Our experiments are performed using a set of representative kernels and several popular scientific benchmarks. Lastly, we quantify a number of power and performance measurements and draw observations and conclusions that can be viewed as a roadmap to achieving energy efficiency in the design and execution of scientific algorithms.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗