Optimizing Threshold for Extreme Scale Analysis (poster).
Abstract not provided.
Moreland, Kenneth D.↗
SEARCH · Engineering Papers
Results for “Scale analysis”
Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.
Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.
Approaching Production In Situ Visualization for Extreme Scale Analysis (SIAM PP Minisymposium).
Abstract not provided.
Moreland, Kenneth D.↗
Hydrate-Bearing Core Characterization: Pore to Core Scale Analysis
The presentation provides a description of NETL's improved capabilities for hydrate-bearing pressure core characterization and visualization.
Pohl, Mathias↗
Development of a Statistical Sampling Method for Uncertainty Analysis with SCALE
A new statistical sampling sequence called Sampler has been developed for the SCALE code system. Random values for the input multigroup cross sections are determined by using the XSUSA program to sample uncertainty data provided in the SCALE covariance library. Using these samples, Sampler computes perturbed self-shielded cross sections and propagates the perturbed nuclear data through any specified SCALE analysis sequence, including those for criticality safety, lattice physics with depletion, and shielding calculations. Statistical analysis of the output distributions provides uncertainties and correlations in the desired responses.
22 GENERAL STUDIES OF NUCLEAR REACTORS↗
Development of a statistical sampling method for uncertainty analysis with SCALE
A new statistical sampling sequence called Sampler has been developed for the SCALE code system. Random values for the input multigroup cross sections are determined by using the XSUSA program to sample uncertainty data provided in the SCALE covariance library. Using these samples, Sampler computes perturbed self-shielded cross sections and propagates the perturbed nuclear data through any specified SCALE analysis sequence, including those for criticality safety, lattice physics with depletion, and shielding calculations. Statistical analysis of the output distributions provides uncertainties and correlations in the desired responses. (authors)
22 GENERAL STUDIES OF NUCLEAR REACTORS↗
Intuitive Visualizations Through Multi-Domain Projections for Performance Analysis at Scale
Performance analysis of parallel scientific codes is becoming increasingly difficult due to the rapidly growing complexity of applications and architectures. Existing tools fall short in providing intuitive views to reveal the root causes of performance problems. We have developed a new paradigm of projecting and visualizing performance data obtained from one domain onto other domains for faster, more intuitive analysis of applications. We gather performance data in three domains: hardware, application, and communication. For each domain, we define projections that allow the data to be viewed in the other domains. This framework is called the HAC model. The model not only allows us to directly compare the data across domains, but also allows us to use data visualization and analysis tools available in the other domains. Using the methodologies and visualization techniques mentioned above, we have demonstrated the careful unscrambling of otherwise tangled measurements caused by adaptive systems. By attributing performance measurements directly to their causes, we are able to visualize performance measurements in the domains most intuitive to the user, which are not necessarily those in which measurements are collected.
97 MATHEMATICS AND COMPUTING↗
Multi-Scale Thermal Analysis for Design of SiC-Based Medium Voltage Motor Drive: Preprint
This paper describes a multi-scale thermal analysis approach for the design of an air-cooled 1.7-kV SiC MOSFET-based medium-voltage variable-speed motor drive. The National Renewable Energy Laboratory supported the project objectives by providing thermal management design support to the project partner The Ohio State University (OSU). The scope of the models and limited resources required efficient and flexible thermal models to be developed. Two modeling techniques are described that significantly reduced model run time and enabled more complex models to be run faster while retaining needed accuracy. The first technique uses the effectiveness-NTU method to extract convection boundary conditions from a computational fluid dynamics (CFD) model that can be applied to a fast-running finite element analysis model. The second is a porous media technique that enables system-level CFD simulations that incorporate effects from heat exchangers (e.g., pin fin heat sinks) that run in a fraction of the time required for fully resolved CFD simulations. The multi-scale approach to the thermal analysis enabled fast and accurate simulation for the converter design ranging from the die level up to the full system with 36 submodules. The modeling results were validated against experimental data from system tests performed by OSU.
DIRECT ENERGY CONVERSION↗
PWR core and spent fuel pool analysis using scale and nestle
The SCALE nuclear analysis code system [SCALE, 2011], developed and maintained at Oak Ridge National Laboratory (ORNL) is widely recognized as high quality software for analyzing nuclear systems. The SCALE code system is composed of several validated computer codes and methods with standard control sequences, such as the TRITON/NEWT lattice physics sequence, which supplies dependable and accurate analyses for industry, regulators, and academia. Although TRITON generates energy-collapsed and space-homogenized few group cross sections, SCALE does not include a full-core nodal neutron diffusion simulation module within. However, in the past few years, the open-source NESTLE core simulator [NESTLE, 2003], originally developed at North Carolina State Univ. (NCSU), has been updated and upgraded via collaboration between ORNL and the Univ. of Tennessee (UT), so it now has a growingly seamless coupling to the TRITON/NEWT lattice physics [Galloway, 2010]. This study presents the methodology used to couple lattice physics data between TRITON and NESTLE in order to perform a three-dimensional full-core analysis employing a 'real-life' Duke Energy PWR as the test bed. The focus for this step was to compare the key parameters of core reactivity and radial power distribution versus plant data. Following the core analysis, following a three cycle burn, a spent fuel pool analysis was done using information generated from NESTLE for the discharged bundles and was compared to Duke Energy spent fuel pool models. The KENO control module from SCALE was employed for this latter stage of the project. (authors)
21 SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLANTS↗
Scalable, In-situ Data Clustering Data Analysis for Extreme Scale Scientific Computing (Final Report)
The objective of this project is to address challenges in the design and development of scalable in-situ data clustering and analytics algorithms and software. Our goal is to develop parallel software consisting of a set of spatio-temporal data clustering and anomaly detection functions, both of which are very important for large-scale analysis and have wide applicability for in-situ runs as well as post-processing analysis. Our design principles for in-situ analysis consider the following: (1) identify parts of the computation can be done close to the data within the nodes, while it is still in memory; (2) extract analysis components can (and should) be performed in remote staging and analysis nodes; (3) develop error-bound approximation methods for applications tolerable for small errors; (4) identify the type of derived distributions and statistics, for spatio-temporal data, that can be kept locally in order to both accelerate computations and meet energy constraints in subsequent iterations and phases; (5) use a self-describing data format so that data can be consistent and understood among local storage (memory and SSDs) and at staging and analysis nodes, thereby providing portability and flexibility; (6) develop service-oriented functions that can schedule in-situ and post-hoc analysis tasks based on the dynamic requirements of applications. Our development focus is to produce the parallel data analysis software/library that will be scalable, reusable, extensible, and generic for applications in different disciplines. The software will be able to run in-situ with the simulations as well as post-hoc analysis. This approach will satisfy many synergistic requirements for data intensive applications executed on data coming from instruments and experiments. In particular, the proposed multilevel approach is directly applicable to perform design tradeoffs for running part of the algorithms near the instruments and the rest on remote (analysis) systems.
97 MATHEMATICS AND COMPUTING↗
Combining In-situ and In-transit Processing to Enable Extreme-Scale Scientific Analysis
With the onset of extreme-scale computing, I/O constraints make it increasingly difficult for scientists to save a sufficient amount of raw simulation data to persistent storage. One potential solution is to change the data analysis pipeline from a post-process centric to a concurrent approach based on either in-situ or in-transit processing. In this context computations are considered in-situ if they utilize the primary compute resources, while in-transit processing refers to offloading computations toa set of secondary resources using asynchronous data transfers. In this paper we explore the design and implementation of three common analysis techniques typically performed on large-scale scientific simulations: topological analysis, descriptive statistics, and visualization. We summarize algorithmic developments, describe a resource scheduling system to coordinate the execution ofvarious analysis workflows, and discuss our implementation using the DataSpaces and ADIOS frameworks that support efficient data movement between in-situ and in-transit computations. Wedemonstrate the efficiency of our lightweight, flexible framework by deploying it on the Jaguar XK6 to analyze data generated by S3D, a massively parallel turbulent combustion code. Our framework allows scientists dealing with the data deluge at extreme scale to perform analyses at increased temporal resolutions, mitigate I/O costs, and significantly improve the time to insight.
Bennett, Janine C.↗
Towards Scaling Law Analysis For Spatiotemporal Weather Data
Compute-optimal scaling laws are relatively well studied for NLP and CV, where objectives are typically single-step and targets are comparatively homogeneous. Weather forecasting is harder to characterize in the same framework: autoregressive rollouts compound errors over long horizons, outputs couple many physical channels with disparate scales and predictability, and globally pooled test metrics can disagree sharply with per-channel, late-lead behavior implied by short-horizon training. We extend neural scaling analysis for autoregressive weather forecasting from single-step training loss to long rollouts and per-channel metrics. We quantify (1) how prediction error is distributed across channels and how its growth rate evolves with forecast horizon, (2) if power law scaling holds for test error, relative to rollout length when error is pooled globally, and (3) how that fit varies jointly with horizon and channel for parameter, data, and compute-based scaling axes. We find strong cross-channel and cross-horizon heterogeneity: pooled scaling can look favorable while many channels degrade at late leads. We discuss implications for weighted objectives, horizon-aware curricula, and resource allocation across outputs.
Kiefer Jr, Alexander [ORNL] (ORCID:000000025398874↗
Design and implementation of I/O performance prediction scheme on HPC systems through large-scale log analysis
Abstract Large-scale high performance computing (HPC) systems typically consist of many thousands of CPUs and storage units used by hundreds to thousands of users simultaneously. Applications from large numbers of users have diverse characteristics, such as varying computation, communication, memory, and I/O intensity. A good understanding of the performance characteristics of each user application is important for job scheduling and resource provisioning. Among these performance characteristics, I/O performance is becoming increasingly important as data sizes rapidly increase and large-scale applications, such as simulation and model training, are widely adopted. However, predicting I/O performance is difficult because I/O systems are shared among all users and involve many layers of software and hardware stack, including the application, network interconnect, operating system, file system, and storage devices. Furthermore, updates to these layers and changes in system management policy can significantly alter the I/O behavior of applications and the entire system. To improve the prediction of the I/O performance on HPC systems, we propose integrating information from several different system logs and developing a regression-based approach to predict the I/O performance. Our proposed scheme can dynamically select the most relevant features from the log entries using various feature selection algorithms and scoring functions, and can automatically select the regression algorithm with the best accuracy for the prediction task. The evaluation results show that our proposed scheme can predict the write performance with up to 90% prediction accuracy and the read performance with up to 99% prediction accuracy using the real logs from the Cori supercomputer system at NERSC.
97 MATHEMATICS AND COMPUTING↗
Scaling patch analysis of planar turbulent mixing layers
Proper scales for the mean flow and Reynolds shear stress in planar turbulent mixing layers are determined from a scaling patch analysis of the mean continuity and momentum equations. By seeking an admissible scaling of the mean continuity equation, a proper scale for the mean transverse flow is determined as V ref =(dδ/dx)U ref , where dδ/dx is the growth rate of the mixing layer width and U ref =U h –U l is the difference between the velocity of the high speed stream U h and the velocity of the low speed stream U l . By seeking an admissible scaling for the mean momentum equation, a proper scale for the kinematic Reynolds shear stress is determined as R uv,ref =U avg V ref =[$\frac{1}{2A_u}$$\frac{dδ}{dx}$]$U^{2}_{ref}$ where A u $_{=}^{def}$(U h –U l )/(U h +U l ) is the normalized velocity difference that emerges naturally in the admissible scaling of the mean momentum equation. Self-similar equations for the scaled mean transverse flow V* and Reynolds shear stress $R^{*}_{uv}$=R uv /R uv,ref are derived from the mean continuity and mean momentum equations. Finally, approximate equations for V* and $R^{*}_{uv}$ are developed and found to agree well with experimental data.
71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
Scaling patch analysis of turbulent planar plume
Proper scaling in turbulent planar plumes is investigated in this study using a scaling patch approach. Based on the scaled boundary conditions, a proper velocity scale for the mean axial flow is the plume centerline velocity U ref = U ctr , and a proper temperature scale for the temperature excess is Θ ref = T ctr – T ∞ , where T ctr is the plume centerline temperature and T ∞ is the ambient fluid temperature. By seeking an admissible scaling, a key concept in the scaling patch approach, for the mean continuity, mean momentum, and mean energy equations, respectively, the following is found: (1) a proper scale for the mean transverse flow is V ref = ( d δ / d x ) U ctr , where d δ / d x is the growth rate of the plume width. (2) A proper scale for the Reynolds shear stress is R vu , ref = U ctr V ref = ( d δ / d x ) U ctr 2 , a mix of the scales for the mean axial and transverse flows. (3) A proper scale for the turbulent heat flux is R v θ , ref = V ref Θ ctr , a mix of the scales for the mean transverse flow and mean temperature excess. The mean transverse flow thus plays a critical role in the scaling of turbulent planar plumes. Approximate functions are developed for the scaled mean transverse flow, Reynolds shear stress, and turbulent temperature flux, and are found to agree favorably with experimental and numerical simulation data. The integral analysis of the mean momentum equation yields a Richardson number Ri, which remains invariant in the axial direction. The Richardson number is defined as Ri = def g β Θ ctr δ t / ( U ctr V ref ) ≈ 1 / 2 , where g is the gravitational acceleration, β is the thermal expansion coefficient, and δ t is the plume half-width based on the mean temperature profile. This Richardson number arises directly from the scaling patch analysis of the mean momentum equation, including both the streamwise and transverse velocity scales.
42 ENGINEERING↗
DynaM: Dynamic Multiresolution Data Representation for Large-Scale Scientific Analysis
Fast growing large-scale systems enable scientific applications to run at a much larger scale and accordingly produce gigantic volumes of simulation output. Such data imposes a grand challenge to post-processing tasks such as visualization and data analysis, because these tasks are often performed at a host machine that is remotely located and equipped with much less memory and storage resources. During the simulation runs, it is also desirable for scientists to be able to interactively monitor and steer the progress of simulation. This requires scientific data to be represented in an efficient form for initial exploration and computation steering. In this paper, we propose DynaM a software framework that can represent scientific data in a multiresolution form, and dynamically organize data blocks into an optimized layout for efficient scientific analysis. DynaM supports a convolution-based multiresolution data representation for abstracting scientific data for visualization at a wide spectrum of resolution. To support the efficient generation and retrieval of different data granularities from such representation, a dynamic data organization in DynaM is enabled to cater distinct peculiarities of different size data blocks for efficient and balanced I/O performance. Our experimental results demonstrate that DynaM can efficiently represent large scientific dataset and speed up the visualization of multidimensional scientific data. An up to 29 times speedup is achieved on Jaguar supercomputer at Oak Ridge National Laboratory.
Tian, Yuan↗