Search NASA⌕ Search

SEARCH · Search NASA

Results for “high performance space 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 307 records · Page 17
All sourcesNASA NTRSDOE OSTI
DOE OSTI2020

Gaussian Process Optimization of Sensitivity-Based Similarity Metrics between New Nuclear Applications and New/Existing Benchmarks

Confidently designing safe, new nuclear criticality experiments requires expert judgement, which could take years of experience. Sensitivity/uncertainty (S/U) analysis can be utilized by less experienced individuals to conservatively estimate uncertainties in important parameters, such as k eff , in newly proposed nuclear applications. This type of analysis relies on matching new nuclear applications with existing benchmark experiments. The Whisper-1.1 software package included in MCNP6.2 ®1 contains more than 1,100 International Criticality Safety Benchmark Evaluation Project (ICSBEP) benchmarks. These benchmarks however rarely match new nuclear applications. The number of benchmarks available to match a given set of materials or geometric configurations varies significantly. Furthermore, recently performed benchmark experiments may not have had enough time to be properly documented and published. Benchmarks are vital for determining the accuracy of nuclear data and can assist nuclear physics and evaluators in improving nuclear data libraries. Exhaustively exploring the parameter space using simulations with software such as MCNP is too computationally expensive. In this work, Gaussian process optimization was implemented to reduce the number of simulations needed for optimization over multiple parameters. This optimization scheme was designed to selectively generate new benchmarks with high sensitivity-based similarity metrics to user-defined nuclear applications. Two benchmark models of spherically nested shells containing plutonium, uranium, tantalum, and water were used in the optimization to match an application containing plutonium plates, stacked in a tantalum reflector, surrounded by water.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗
DOE OSTI2023

Second Report of the Nuclear Data Subcommittee of the Nuclear Science Advisory Committee

The central importance of the nuclear data curated by the US Nuclear Data Program (USNDP) for clean energy generation, national security, nonproliferation, medical applications, and space exploration as well as basic science was described in a prior report issued by the DOE/NSF Nuclear Science Advisory Committee subcommittee on Nuclear Data (NSAC-ND) in September 2022. In this report, we present a set of fourteen (14) recommendations that will enhance and advance DOE-NP's stewardship of nuclear data. The first three recommendations focus on the existing core USNDP capabilities, namely: 1) Support the nuclear structure evaluation workforce to improve the currency, consistency, and accessibility of the Evaluated Nuclear Structure Data File (ENSDF); 2) Enhance nuclear reaction evaluation within the USNDP in support of the Evaluated Nuclear Data File (ENDF) through expansion of the workforce and integration of high-performance computing, automation, and machine learning and; 3) Continue atomic mass evaluation in support AME and NUBASE databases. This is followed by eight (8) recommendations representing new cross-cutting initiatives involving both measurement and evaluation to address outstanding nuclear data needs. These new initiatives require a highly trained, diverse workforce that includes personnel with expertise from both inside and outside the nuclear physics community from which evaluators have traditionally been recruited. As such, many of these initiatives are accomplished via a Topical Nuclear Data Collaborations (TNDC). A TNDC is made up of domestic and international stakeholders, subject matter and nuclear data experts, and nuclear data evaluators and features a workforce development plan to ensure that nuclear data evaluators maintain currency in the relevant applications and are seen as equity partners in the endeavor. These include: 1) Establish a coordinated effort to improve evaluation and modeling in nuclear astrophysics for stellar dynamics, multi-messenger astronomy and nucleosynthesis; 2) Initiate a TNDC to develop and maintain nuclear structure evaluation beyond discrete states, including nuclear level densities, photon strength functions and photonuclear data for improved reaction modeling, and exploring nuclear structure at finite temperature; 3) Create a TNDC to perform correlated fission data evaluation, including cross sections, fragment yields, v(A), v(E n ) for nuclear energy, national security, nonproliferation and basic science; 4) From a panel of subject matter experts to establish and annually update a roster of key decay data to nurture its accelerated dissemination including both measurement and evaluation for targeted high-value nuclides for national security, nonproliferation and medical applications; 5) Comprehensive, consistent neutron-induced structure and reaction data for nuclear energy, national security, nonproliferation and planetary nuclear spectroscopy; 6) Charged-particle stopping powers for detector design, space effects and ion beam therapy; 7) High-energy reactions for space exploration and medical nuclide production, and; 8) The creation of an infrastructure for open data and data preservation for use by the entire nuclear physics community. All told, these initiatives require approximately $6.5M increase in NP support of the USNDP in fiscal year 2023 dollars and would require at least 3-5 years to carry out due to the length of time needed to recruit and train new nuclear data researchers. This relatively modest investment would help ensure that the fruits of the nuclear data research carried out by DOE-NP and its collaborators would be brought to bear to address some of the most important needs of our nation and the world. To ensure effective execution of this plan, we present an overview of recruitment, training, and retention goals for the USNDP, the centerpiece of which is a mutually agreed upon code of conduct. Finally, we identify the facility and instrumentation needed to perform the recommended experimental activities. This includes a short review of target fabrication capabilities, reactors, neutron beam, light- and heavy-stable ion, gamma-ray, high-energy and radioactive ion beam facilities. Lastly, a more complete appendix of experimental facilities previously compiled is included with new input provided for 6 facilities.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗
DOE OSTI2023

A MultiGPU Performance-Portable Solution for Array Programming Based on Kokkos

Today, multiGPU nodes are widely used in high-performance computing and data centers. However, current programming models do not provide simple, transparent, and portable support for automatically targeting multiple GPUs within a node on application areas of array programming. In this paper, we describe a new application programming interface based on the Kokkos programming model to enable array computation on multiple GPUs in a transparent and portable way across both NVIDIA and AMD GPUs. We implement different variations of this technique to accommodate the exchange of stencils (array boundaries) among different GPU memory spaces, and we provide autotuning to select the proper number of GPUs, depending on the computational cost of the operations to be computed on arrays, that is completely transparent to the programmer. We evaluate our multiGPU extension on Summit (#5 TOP500), with six NVIDIA V100 Volta GPUs per node, and Crusher that contains identical hardware/software as Frontier (#1 TOP500), with four AMD MI250X GPUs, each with 2 Graphics Compute Dies (GCDs)for a total of 8 GCDs per node. We also compare the performance of this solution against the use of MPI + Kokkos, which is the cur-rent de facto solution for multiple GPUs in Kokkos. Our evaluation shows that the new Kokkos solution provides good scalability for many GPUs and a faster and simpler solution (from a programming productivity perspective) than MPI + Kokkos.

Valero Lara, Pedro↗
DOE OSTI2020

Simulation-driven optimization of high-order meshes in ALE hydrodynamics

Here we propose tools for high-order mesh optimization and demonstrate their benefits in the context of multi-material Arbitrary Lagrangian-Eulerian (ALE) compressible shock hydrodynamic applications. The mesh optimization process is driven by information provided by the simulation which uses the optimized mesh, such as shock positions, material regions, known error estimates, etc. These simulation features are usually represented discretely, for instance, as finite element functions on the Lagrangian mesh. The discrete nature of the input is critical for the practical applicability of the algorithms we propose and distinguishes this work from approaches that strictly require analytical information. Our methods are based on node movement through a high-order extension of the Target-Matrix Optimization Paradigm (TMOP). The proposed formulation is fully algebraic and relies only on local Jacobian matrices, so it is applicable to all types of mesh elements, in 2D and 3D, and any order of the mesh. We discuss the notions of constructing adaptive target matrices and obtaining their derivatives, reconstructing discrete data in intermediate meshes, node limiting that enables improvement of global mesh quality while preserving space-dependent local mesh features, and appropriate normalization of the objective function. The adaptivity methods are combined with automatic ALE triggers that can provide robustness of the mesh evolution and avoid excessive remap procedures. The benefits of the new high-order TMOP technology are illustrated on several simulations performed in the high-order ALE application BLAST.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2024

Approaches for the Simulation of Coupled Processes in Evolving Fractured Porous Media Enabled by Exascale Computing

Models have historically represented fractured porous media with continuum descriptions that characterize the media using bulk parameters. The impact of small-scale features is not captured in these models, although they may be controlling the performance of subsurface applications. Pore-scale models can simulate processes in small-scale features by representing the pore space geometry explicitly but are computationally expensive for large domains. The alternative multiscale approach entails the combination of pore-scale and continuum-scale descriptions in a single framework. We use Chombo-Crunch, a computational capability that discretizes complex geometries with an adaptive, embedded boundary method to contrast these two approaches. Chombo-Crunch takes advantage of recent computational performance and memory bandwidth improvements resulting from the emergence of exascale computing resources. These combined improvements enable the efficient simulation of reactive transport in fractured media with a high degree of fidelity and the ability to capture the control small-scale processes exert on the overall medium evolution.

42 ENGINEERING↗
DOE OSTI2020

Basic factors for acquiring, correcting, and interpreting probe current-voltage characteristic in moderate-collisional plasma for determining electron energy distribution

Methods for inferring the electron distribution function (EDF) from a probe current–voltage trace are reviewed for moderate-collisional plasma. The explicit accounting of particle collisions, modeling sheath depletion, and electron refilling processes are rarely performed when analyzing probe measurements, yet they are critical for regime validity, EDF-profile accuracy, and space potential quantification precision. Furthermore, the advantages, limitations, and utilization of large-area wall probes in moderate-collisional plasma, capable of highly resolved indications of energetic electrons in plasma, are outlined.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2020

Finite-temperature coupled cluster: Efficient implementation and application to prototypical systems

In this work, we discuss the theory and implementation of the finite temperature coupled cluster singles and doubles (FT-CCSD) method including the equations necessary for an efficient implementation of response properties. Numerical aspects of the method including the truncation of the orbital space and integration of the amplitude equations are tested on some simple systems, and we provide some guidelines for applying the method in practice. The method is then applied to the 1D Hubbard model, the uniform electron gas (UEG) at warm, dense conditions, and some simple materials. The performance of model systems at high temperatures is encouraging: for the one-dimensional Hubbard model, FT-CCSD provides a qualitatively accurate description of finite-temperature correlation effects even at U = 8, and it allows for the computation of systematically improvable exchange–correlation energies of the warm, dense UEG over a wide range of conditions. We highlight the obstacles that remain in using the method for realistic ab initio calculations on materials

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2023

Sparse Symmetric Format for Tucker Decomposition

Tensor-based methods are receiving renewed attention in recent years due to their prevalence in diverse real-world applications. There is considerable literature on tensor representations and algorithms for tensor decompositions, both for dense and sparse tensors. Many applications in hypergraph analytics, machine learning, psychometry, and signal processing result in tensors that are both sparse and symmetric, making them an important class for further study. Similar to the critical Tensor Times Matrix chain operation (TTM c ) in general sparse tensors, the $\underline{S}$ parse $\underline{S}$ ymmetric $\underline{T}$ ensor $\underline{T}$ imes $\underline{S}$ ame $\underline{M}$ atrix $\underline{c}$ hain (S 3 TTM c ) operation is compute and memory intensive due to high tensor order and the associated factorial explosion in the number of non-zeros. We present the novel Compressed Sparse Symmetric (CSS) format for sparse symmetric tensors, along with an efficient parallel algorithm for the S 3 TTM c operation. We theoretically establish that S 3 TTM c on CSS achieves a better memory versus run-time trade-off compared to state-of-the-art implementations, and visualize the variation of the performance gap over the parameter space. We demonstrate experimental findings that confirm these results and achieve up to 2.72× speedup on synthetic and real datasets. The scaling of the algorithm on different test architectures is also showcased to highlight the effect of machine characteristics on algorithm performance.

42 ENGINEERING↗
DOE OSTI2024

Multiscale and Machine Learning Modeling for Process-informed Microstructure Prediction in Additively Manufactured Materials Using MALAMUTE

Advanced Materials and Manufacturing Technologies (AMMT) program under the Department of Energy Office of Nuclear Energy, aims to develop and qualify additively-manufactured materials for nuclear applications. The key challenges to these efforts are the microstructural variabilities observed on the AM products and their impact on the properties and performance of the material in extreme environments. AMMT is using a combination of high-through-put experimental and modeling techniques to accelerate the qualification efforts. Conventionally, in-situ and ex-situ characterizations and testing are performed to correlate different aspects of the AM process to the final product and its performance. However, adopting a trial-and-error approach to experimentally evaluate the vast range of process parameters required to capture the microstructural variabilities is cost-prohibitive. Modeling and simulation provide a comparatively inexpensive way to understand and correlate the microstructural evolution to the processing conditions. The modeling and simulation work-packages within the AMMT program aims to use physics-based and machine learning modeling capabilities to develop a digital twin for AM that can correlate the process conditions to the final product and establish a process-structure-property-performance (PSPP) correlation for AM materials. The melting and subsequent solidification that occurs during the AM process is a complex phenomenon that requires multiscale multiphysics analysis. Idaho National Laboratory’s (INL) Multiphysics Object-Oriented Simulation Environment (MOOSE), specifically the MOOSE Application Library for Advanced Manufacturing UTilitiEs (MALAMUTE) software, provides an ideal platform for developing the multiphysics multiscale model to explore the intricacies of the microstructural evolution during the AM processes within a single framework. Furthermore, given that such full-fidelity simulations can be computationally intensive, reduced order models are necessary to explore the PSPP space for AM materials in an efficient, reliable, and cost-effective way. This work package focuses on understanding the role of process variabilities on the various microstructural characteristics of the AM materials. Microstructures unique to AM materials, such as compositional micro-heterogeneity and dislocation cells, are of particular interest here since they can influence the creep properties and radiation performance. In fiscal year (FY) 24, we significantly advanced upon our work in the last fiscal year, both on physics-based and ML models. The alloy solidification model available in MOOSE has been extended to incorporate the thermodynamic properties and free energy relevant to 316SS. The model demonstrates the Cr segregation that occurs during solidifcation. It is demonstrated that rate of solidification and solute segregation is primarily influence by the cooling rate dictating the level of freezing. This work captures the microstructural variabilities at the subgrain level that are often missing in the part-scale models. With an aim to connect the microstructural evolution model to realistic process conditions, a reduced order model is developed for predicting the thermal conditions around meltpool from high-fidelity process simulations. Furthermore, machine learning approach is used to accelerate the temperature prediction during the AM process. In the following years, MALAMUTE will be used to connect different aspects of the models and quantitatively predict the microstructural evolution. The developed ML-based surrogate model will consider the process conditions as the input to predict the microstructural features in a cost-effective way. The generated microstructures can be used by other work packages under AMMT to evaluate the properties and environmental response of the material at the mesoscale. Thus, this work help identify the key microstructural features at the subgrain level that are significant in property/performance prediction of the AM products. This work will provide inputs to the large-scale process variability models to reevaluate and validate assumptions/simplifications made in the part-scale models. Furthermore, through active learning this work will help identify the data need from both modeling and experimental sides for development of a robust digital twin for AM.

36 MATERIALS SCIENCE↗
DOE OSTI2021

Dotted-line FLEET for two-component velocimetry

Femtosecond laser electronic excitation tagging (FLEET) is a powerful unseeded velocimetry technique typically used to measure one component of velocity along a line, or two or three components from a dot. In this Letter, we demonstrate a dotted-line FLEET technique which combines the dense profile capability of a line with the ability to perform two-component velocimetry with a single camera on a dot. Our set-up uses a single beam path to create multiple simultaneous spots, more than previously achieved in other FLEET spot configurations. We perform dotted-line FLEET measurements downstream of a highly turbulent, supersonic nitrogen free jet. Dotted-line FLEET is created by focusing light transmitted by a periodic mask with rectangular slits of 1.6 × 40 mm 2 and an edge-to-edge spacing of 0.5 mm, then focusing the imaged light at the measurement region. Up to seven symmetric dots spaced approximately 0.9 mm apart, with mean full-width at half maximum diameters between 150 and 350 µm, are simultaneously imaged. Both streamwise and radial velocities are computed and presented in this Letter.

47 OTHER INSTRUMENTATION↗
DOE OSTI2023

Denoising diffusion algorithm for inverse design of microstructures with fine-tuned nonlinear material properties

Here we introduce a denoising diffusion algorithm to discover microstructures with nonlinear fine-tuned properties. Denoising diffusion probabilistic models are generative models that use diffusion-based dynamics to gradually denoise images and generate realistic synthetic samples. By learning the reverse of a Markov diffusion process, we design an artificial intelligence to efficiently manipulate the topology of microstructures to generate a massive number of prototypes that exhibit constitutive responses sufficiently close to designated nonlinear constitutive behaviors. To identify the subset of microcstructures with sufficiently precise fine-tuned properties, a convolutional neural network surrogate is trained to replace high-fidelity finite element simulations to filter out prototypes outside the admissible range. Results of this study indicate that the denoising diffusion process is capable of creating microstructures of fine-tuned nonlinear material properties within the latent space of the training data. More importantly, this denoising diffusion algorithm can be easily extended to incorporate additional topological and geometric modifications by introducing high-dimensional structures embedded in the latent space. Numerical experiments are conducted on the open-source mechanical MNIST data set (Lejeune, 2020). Consequently, this algorithm is not only capable of performing inverse design of nonlinear effective media, but also learns the nonlinear structure–property map to quantitatively understand the multiscale interplay among the geometry, topology, and their effective macroscopic properties.

42 ENGINEERING↗
DOE OSTI2023

Automated Air Sealing Demonstration: Denver Federal Center Building 40

This project aimed to demonstrate building airtightness achieved by automated air sealing in a commercial building setting. The automated air sealing demonstrated in this project uses a modified blower door to pressurize and distribute the sealing aerosol to achieve the required building airtightness. To demonstrate this technology, Building 40 at the Denver Federal Center, a federally owned campus under the jurisdiction, custody and control of the U.S. General Services Administration (GSA), was selected for testing (Figure 2). This building is currently undergoing several retrofit projects, including increasing insulation, installing high-performance windows, and adding advanced equipment, control systems, and automated air sealing. This demonstration project involved installing automated air sealing and measuring the improvements in the building's airtightness. The automated air sealing was installed over two days by AeroBarrier, the vendor. An external blower door test contractor measured the airtightness of the demonstration space before and after air sealing. The new airtightness value and the percentage increase in airtightness were used to evaluate the energy savings potential of automated air sealing. These airtightness values were used to compute the energy savings and CO 2 emissions reduction for different climate zones, building types, and initial airtightness conditions. In addition, the heating, ventilation and air conditioning (HVAC) load reduction attributable to the reduced building air leakage was studied. This study included research to determine the cost and time reduction of automated air sealing. Finally, the automated air sealing performance was evaluated qualitatively using a focus group discussion that included GSA and Bristol, the general contractor. The installation has shown the demonstration site, with a floor size of 4,462 ft 2 , air leakage has reduced by more than 50% in less than 7 hours, including preparation, site sealing, and cleanup. The performance objectives were classified as quantitative or qualitative based on the evaluation metrics used to assess the project’s success. The key performance objectives for this project were the level of airtightness achieved, the time and cost required to perform the sealing, and the HVAC load reduction attributable to air sealing. Table 1 shows the quantitative performance objectives.

42 ENGINEERING↗
DOE OSTI2025

Inertial Confinement Fusion Design Search Using Bayesian Optimization

Inertial confinement fusion (ICF) experiments rely on complex multi-physics simulation codes such as the Lawrence Livermore National Laboratory-developed HYDRA to guide design work. However, these simulations have several dozen tunable parameters and can be computationally expensive. This makes searching the parameter space challenging and time-consuming. Recently developed automated tools utilize Bayesian optimization to search these high-dimensional parameter spaces for optimal designs. The optimization tools run 2D integrated simulations in HYDRA to converge on a design that produces specified scalar or vector outputs. In this paper, we apply the Bayesian optimization tools to two common tuning scenarios. First, we tune simulation inputs to match measurements of a well-characterized experiment at the National Ignition Facility. This type of tuning is commonly performed to compensate for the use of simplified simulation settings (e.g. reduced resolution) or to account for missing physics in the simulations. Second, we search for an ICF simulation design that has a particular radiation drive profile. These optimizations replicate the kinds of tuning researchers routinely perform, but do so with significantly reduced manual effort. This approach demonstrates a powerful and efficient pathway toward autonomous, high-fidelity design optimization for future ICF experiments.

Bayesian optimization↗
DOE OSTI2025

Exploring transport-enabled gates with integrated optical addressing to demonstrate high fidelity control of trapped ion qubits in a scalable quantum computer

In recent years, experiments involving micofabricated surface ion traps have grown in complexity, and as this complexity grows, a common design has emerged in the form of quantum charge-coupled device architecture. This architecture, already utilized on multiple systems, supports multiple zones on a device for areas such as memory or computation. The shutting process between these zones is a process often seen to be minimized. An additional component to scalable surface trapped-ion experiments includes some form of integrated photonics, as free space lasers have difficulty scaling to many ions. Here, we discuss recent results in utilizing the shuttling process with integrated photonics to perform a specific type of gate, as well as demonstrating state preparation. Furthermore, we show that these gates can be utilized as a type of optical modulator as the Doppler shift that occurs during shuttling will make the light seen by shuttled ions different than that seen by stationary ions. Furthermore, we show that the shuttling operation can be utilized as an optical modulator, where the Doppler shift changes the frequency from that of a stationary ion.

42 ENGINEERING↗
DOE OSTI2022

Impact of large-scale effects on mass transfer and concentration polarization in Reverse Osmosis membrane systems

Here, we present well-resolved computational fluid dynamics simulations of a large-scale reverse osmosis membrane-spacer configuration (1 m). Our computational model solves the flow and transport equations with variable solute-dependent properties. We utilize a high resolution computational mesh to resolve all relevant length scales associated with spacer-induced mixing and thin concentration boundary layers. An important contribution of this work is the development of a modified mass-transfer correlation that accounts for the development of the concentration boundary layer along the channel. A set of 2D axisymmetric simulations were performed for a spiral wound module layer with varying cross-flow conditions and spacer diameters which indicate a significant entrance length effect for concentration profile development at lower flow rates while mixing effects dominate at higher flow rates. The mass-transfer correlations at higher flow rates compare well with published correlations while a surrogate model for Sherwood number was obtained that depends on an additional similarity variable that accounted for entrance length effects at lower flow rates. Finally, a large-scale membrane-spacer design relevant to high-pressure reverse osmosis is studied with a non-uniform arrangement of spacers, which indicate a substantial saving in pressure drop (~40%) compared to traditional uniformly spaced pattern with minor variations (~2%) in concentration polarization, product water quality (~1%) and water recovery (~7%) compared to a uniform spacer pattern.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2022

Data augmentation for disruption prediction via robust surrogate models

The goal of this work is to generate large statistically representative data sets to train machine learning models for disruption prediction provided by data from few existing discharges. Such a comprehensive training database is important to achieve satisfying and reliable prediction results in artificial neural network classifiers. Here, we aim for a robust augmentation of the training database for multivariate time series data using Student t process regression. We apply Student t process regression in a state space formulation via Bayesian filtering to tackle challenges imposed by outliers and noise in the training data set and to reduce the computational complexity. Thus, the method can also be used if the time resolution is high. We use an uncorrelated model for each dimension and impose correlations afterwards via colouring transformations. We demonstrate the efficacy of our approach on plasma diagnostics data of three different disruption classes from the DIII-D tokamak. To evaluate if the distribution of the generated data is similar to the training data, we additionally perform statistical analyses using methods from time series analysis, descriptive statistics and classic machine learning clustering algorithms.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2020

Indicator-directed Dynamic Power Management for Iterative Workloads on GPU-Accelerated Systems

Modern high-performance and warehouse computing centers show strong interest in minimizing system power consumption while satisfying customers’ quality of service (QoS). Dynamic voltage and frequency scaling (DVFS) is effective for achieving this goal. Nevertheless, automating the process online and making it transparent to users must address three major challenges: (1) Complexity — today’s hardware components (e.g., CPUs, GPUs, memory, network, etc.) can be configured in several or dozens of frequency/voltage states for satisfying divergent system demands. Given their combination and the emergence of heterogeneity, searching the optimal configuration in the design space online can be timing consuming. (2) QoS guarantee — user-defined objectives such as power constraint and performance target must be monitored, predicted and ensured at the best effort. (3) Adaptability — various known and unknown workloads run on systems. Workloads characteristics should be quickly determined and configurations dynamically adjusted in accord with workloads and QoS. In this work, we focus on applications exhibiting an interesting feature – iterative or periodic, which is common among conventional HPC and emerging machine learning workloads. We propose an online dynamic power-performance (ODPP) management framework to dynamically adjust GPU DVFS configurations to meet performance and power objectives and constraints, without any code annotation or intrusion. Particularly, ODPP extracts the performance and power indicators for applications from their resources utilization profiles in a short episode. It further automatically constructs an accurate model that infers from the indicators how the application's performance and power vary with GPU core and memory frequencies. Aided with the model, for both seen and unseen applications, ODPP can quickly determine the most appropriate DVFS configuration for their execution. We evaluate ODPP on an NVIDIA GPU using multiple exascale computing (ECP) and deep learning applications.

Zou, Pengfei↗
DOE OSTI2022

Data augmentation for disruption prediction via robust surrogate models

The goal of this work is to generate large statistically representative datasets to train machine learning models for disruption prediction provided by data from few existing discharges. Such a comprehensive training database is important to achieve satisfying and reliable prediction results in artificial neural network classifiers. Here, we aim for a robust augmentation of the training database for multivariate time series data using Student-t process regression. We apply Student-t process regression in a state space formulation via Bayesian filtering to tackle challenges imposed by outliers and noise in the training data set and to reduce the computational complexity. Thus, the method can also be used if the time resolution is high. We use an uncorrelated model for each dimension and impose correlations afterwards via coloring transformations. We demonstrate the efficacy of our approach on plasma diagnostics data of three different disruption classes from the DIII-D tokamak. To evaluate if the distribution of the generated data is similar to the training data, we additionally perform statistical analyses using methods from time series analysis, descriptive statistics, and classic machine learning clustering algorithms.

97 MATHEMATICS AND COMPUTING↗