Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “performance tuning”

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 1,045 records · Page 58

Deep learning to estimate permeability using geophysical data

Time-lapse electrical resistivity tomography (ERT) is a popular geophysical method to estimate three-dimensional (3D) permeability fields from electrical potential difference measurements. Traditional inversion and data assimilation methods are used to ingest this ERT data into hydrogeophysical models to estimate permeability. Due to ill-posedness and the curse of dimensionality, existing inversion strategies provide poor estimates and low resolution of the 3D permeability field. Recent advances in deep learning provide us with powerful algorithms to overcome this challenge. This paper presents a deep learning (DL) framework to estimate the 3D subsurface permeability from time-lapse ERT data. To test the feasibility of the proposed framework, we train DL-enabled inverse models on simulation data. Each measurement in both synthetic and field data is standardized by removing the mean and scaling the time-series to unit variance. This pre-processing step is necessary to bring simulation data closer to field observations. Subsurface process models based on hydrogeophysics are used to generate this synthetic data. Training performed on limited simulation data resulted in the DL model over-fitting. An advanced data augmentation based on mixup is implemented to generate additional training samples to overcome this issue. This mixup technique creates weakly labeled (low-fidelity) samples from strongly labeled (high-fidelity) data. The weakly labeled training data is then used to develop DL-enabled inverse models and reduce over-fitting. As both time-lapse ERT (1133048 features/realization) and 3D permeability (585453 features/realization) data samples are from a high-dimensional space, principal component analysis (PCA) is employed to reduce dimensionality. Encoded ERT and encoded permeability are generated using the trained PCA estimators. A deep neural network is then trained to map the encoded ERT to encoded permeability. This mixup training and unsupervised learning allowed us to build a fast and reasonably accurate DL-based inverse model under limited simulation data. Results show that proposed weak supervised learning can capture salient spatial features in the 3D permeability field. Quantitatively, the average mean squared error (in terms of the natural log) on the strongly labeled training, validation, and test datasets is less than 0.5. The R 2 -score (global metric) is greater than 0.75, and the percent error in each cell (local metric) is less than 10%. Finally, an added benefit in terms of computational cost is that the proposed DL-based inverse model is at least O(10 4 ) times faster than running a forward model once it is trained. Data generation, DL model training, and hyperparameter tuning to identify optimal neural network architectures utilized high-performance computing resources while the DL inference is performed on a standard laptop. Approximately, O(10 5 ) processor hours are used for generating data and DL tuning and training. We acknowledge that the data generation and DL model development are expensive. But once a DL model is trained, it can be re-used for inversion rapidly for the given system, with set physics and domain. Note that traditional inversion may require multiple forward model simulations (e.g., in the order of 10 to 1000), which are very expensive. This computational savings ≈ O(10 5 ) – O(10 7 )) makes the proposed DL-based inverse model attractive for subsurface imaging and real-time ERT monitoring applications due to fast and yet reasonably accurate estimations of permeability field.

58 GEOSCIENCES↗

Optimizing Excited-State Electronic-Structure Codes for Intel Knights Landing: A Case Study on the BerkeleyGW Software

We profile and optimize calculations performed with the BerkeleyGW code on the Xeon-Phi architecture. BerkeleyGW depends both on hand-tuned critical kernels as well as on BLAS and FFT libraries. We describe the optimization process and performance improvements achieved. We discuss a layered parallelization strategy to take advantage of vector, thread and node-level parallelism. We discuss locality changes (including the consequence of the lack of L3 cache) and effective use of the on-package high-bandwidth memory. We show preliminary results on Knights-Landing including a roofline study of code performance before and after a number of optimizations. We find that the GW method is particularly well-suited for many-core architectures due to the ability to exploit a large amount of parallelism over plane-wave components, band-pairs, and frequencies.

97 MATHEMATICS AND COMPUTING↗

I/O Bottleneck Detection and Tuning: Connecting the Dots using Interactive Log Analysis

Using parallel file systems efficiently is a tricky problem due to inter-dependencies among multiple layers of I/O software, including high-level I/O libraries (HDF5, netCDF, etc.), MPI-IO, POSIX, and file systems (GPFS, Lustre, etc.). Profiling tools such as Darshan collect traces to help understand the I/O performance behavior. However, there are significant gaps in analyzing the collected traces and then applying tuning options offered by various layers of I/O software. Seeking to connect the dots between I/O bottleneck detection and tuning, we propose DXT Explorer, an interactive log analysis tool. In this paper, we present a case study using our interactive log analysis tool to identify and apply various I/O optimizations. We report an evaluation of performance improvement achieved for four I/O kernels extracted from science applications.

Bez, Jean Luca↗

Upgrade to Fixed and Translating Scintillation-Based Loss Detector System in the Fermilab Drift Tube Linac

The closed-off structure of the Fermilab Drift Tube Linac precludes a robust array of instrumentation from directly monitoring the H- beam that is accelerated from 750 keV to 116 MeV. To improve beam tuning and op-erational assessment of Drift Tube Linac performance, scintillator-based loss monitors were previously installed along the exterior of the first two accelerating cavities to assess low energy beam losses. Here we present a recent upgrade to the loss monitor system, including significant improvements in analog signal processing to address baseline-interfering noise; digitization of the signals to enable regular operational use and tuning; and a new remote operation upgrade of the translating loss monitor with precise positioning of the loss monitor along its nine-foot track. Data from the fixed and translating de-tectors collected under varying beam conditions validate the utility of the upgrade.

Chen, Erin V. [Fermilab] (ORCID:0000000215467899)↗

Performance-Aligned LLMs for Generating Fast HPC Code

Optimizing scientific software is a difficult task because codebases are often large and complex, and performance can depend upon several factors including the algorithm, its implementation, and hardware among others. Causes of poor performance can originate from disparate sources and be difficult to diagnose. Recent years have seen a multitude of work that use large language models (LLMs) to assist in software development tasks. However, these tools are trained to model the distribution of code as text, and are not specifically designed to understand performance aspects of code. In this work, we introduce a reinforcement learning based methodology to align the outputs of code LLMs with performance. This allows us to build upon the current code modeling capabilities of LLMs and extend them to generate better performing code. Here, we demonstrate that our fine-tuned model improves the expected speedup of generated code over base models for a set of benchmark tasks from 0.9 to 1.6 for serial code and 1.9 to 4.5 for OpenMP parallel code.

Computer science↗

Final Focus Test Facility ATF2 Status

ATF2 is a final-focus test beam line which aims to focus the low emittance beam from the ATF damping ring to a vertical size of about 37 nm and to demonstrate nanometre level beam stability. Several advanced beam diagnostics and feedback tools are used. In December 2008, construction and installation were completed and beam commissioning started, supported by an international team of Asian, European and American scientists. In this paper, the present status and performance of the recently deployed ATF2 systems are briefly described, based on the first experience with beam measurements and tuning during winter, spring and early autumn of 2009. The near and longer term plans are outlined as well. The ATF collaboration has completed the construction of ATF2 and has started its commissioning. Important experience operating the new cavity BPM and BSM instrumentation in real conditions has been gained and first beam measurements have been performed in a magnetic configuration with reduced optical demagnification. Both horizontal and vertical emittances were successfully tuned and measured in the extraction line, with values approaching the design values of 2 nm and 12 pm, respectively. First checks of the first order optics along the beam line and at the IP were also done. Hardware developments for the second ATF2 goal are being pursued in parallel with the present commissioning work for the first goal. The collaboration is also preparing several near and long terms plans for ATF2. In the next few years, information very valuable for any future collider with local chromaticity correction and tuning of very low emittance beams can be expected. In the previous experience at the FFTB, the smallest vertical beam sizes which were achieved were about 70 nanometers. The work described here continues to address this largely unexplored regime in a systematic way.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Measurement and analysis of GPU-accelerated applications with HPCToolkit

To address the challenge of performance analysis on the US DOE’s forthcoming exascale supercomputers, Rice University has been extending its HPCToolkit performance tools to support measurement and analysis of GPU-accelerated applications. To help developers understand the performance of accelerated applications as a whole, HPCToolkit’s measurement and analysis tools attribute metrics to calling contexts that span both CPUs and GPUs. To measure GPU-accelerated applications efficiently, HPCToolkit employs a novel wait-free data structure to coordinate monitoring and attribution of GPU performance. To help developers understand the performance of complex GPU code generated from high-level programming models, HPCToolkit constructs sophisticated approximations of call path profiles for GPU computations. To support fine-grained analysis and tuning, HPCToolkit uses PC sampling and instrumentation to measure and attribute GPU performance metrics to source lines, loops, and inlined code. To supplement fine-grained measurements, HPCToolkit can measure GPU kernel executions using hardware performance counters. To provide a view of how an execution evolves over time, HPCToolkit can collect, analyze, and visualize call path traces within and across nodes. Finally, on NVIDIA GPUs, HPCToolkit can derive and attribute a collection of useful performance metrics based on measurements using GPU PC samples. Here, we illustrate HPCToolkit’s new capabilities for analyzing GPU-accelerated applications with several codes developed as part of the Exascale Computing Project.

97 MATHEMATICS AND COMPUTING↗

Shock timing on the National Ignition Facility: First Experiments

An experimental campaign to tune the initial shock compression sequence of capsule implosions on the National Ignition Facility (NIF) was initiated in late 2010. The experiments use a NIF ignition-scale hohlraum and capsule that employs a reentrant cone to provide optical access to the shocks as they propagate in the liquid deuterium-filled capsule interior. The strength and timing of the shock sequence is diagnosed with velocity interferometry that provides target performance data used to set the pulse shape for ignition capsule implosions that follow. From the start, these measurements yielded significant new information on target performance, leading to improvements in the target design. We describe the results and interpretation of the initial tuning experiments.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Advanced Control Considerations for Turbofan Engine Design

This paper covers the application of a model-based engine control (MBEC) methodology featuring a self tuning on-board model for an aircraft turbofan engine simulation. The nonlinear engine model is capable of modeling realistic engine performance, allowing for a verification of the advanced control methodology over a wide range of operating points and life cycle conditions. The on-board model is a piece-wise linear model derived from the nonlinear engine model and updated using an optimal tuner Kalman Filter estimation routine, which enables the on-board model to self-tune to account for engine performance variations. MBEC is used here to show how advanced control architectures can improve efficiency during the design phase of a turbofan engine by reducing conservative operability margins. The operability margins that can be reduced, such as stall margin, can expand the engine design space and offer potential for efficiency improvements. Application of MBEC architecture to a nonlinear engine simulation is shown to reduce the thrust specific fuel consumption by approximately 1% over the baseline design, while maintaining safe operation of the engine across the flight envelope.

control↗

National Ignition Campaign (NIC) Precision Tuning Series Shock Timing Experiments

A series of precision shock timing experiments have been performed on NIF. These experiments continue to adjust the laser pulse shape and employ the adjusted cone fraction (CF) in the picket (1st 2 ns of the laser pulse) as determined from the re-emit experiment series. The NIF ignition laser pulse is precisely shaped and consists of a series of four impulses, which drive a corresponding series of shock waves of increasing strength to accelerate and compress the capsule ablator and fuel layer. To optimize the implosion, they tune not only the strength (or power) but also, to sub-nanosecond accuracy, the timing of the shock waves. In a well-tuned implosion, the shock waves work together to compress and heat the fuel. For the shock timing experiments, a re-entrant cone is inserted through both the hohlraum wall and the capsule ablator allowing a direct optical view of the propagating shocks in the capsule interior using the VISAR (Velocity Interferometer System for Any Reflector) diagnostic from outside the hohlraum. To emulate the DT ice of an ignition capsule, the inside of the cone and the capsule are filled with liquid deuterium.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

An Integrated Approach for Aircraft Engine Performance Estimation and Fault Diagnostics

A Kalman filter-based approach for integrated on-line aircraft engine performance estimation and gas path fault diagnostics is presented. This technique is specifically designed for underdetermined estimation problems where there are more unknown system parameters representing deterioration and faults than available sensor measurements. A previously developed methodology is applied to optimally design a Kalman filter to estimate a vector of tuning parameters, appropriately sized to enable estimation. The estimated tuning parameters can then be transformed into a larger vector of health parameters representing system performance deterioration and fault effects. The results of this study show that basing fault isolation decisions solely on the estimated health parameter vector does not provide ideal results. Furthermore, expanding the number of the health parameters to address additional gas path faults causes a decrease in the estimation accuracy of those health parameters representative of turbomachinery performance deterioration. However, improved fault isolation performance is demonstrated through direct analysis of the estimated tuning parameters produced by the Kalman filter. This was found to provide equivalent or superior accuracy compared to the conventional fault isolation approach based on the analysis of sensed engine outputs, while simplifying online implementation requirements. Results from the application of these techniques to an aircraft engine simulation are presented and discussed.

imon, Donald L.↗

Thermo-structural design of the European DEMO water-cooled blanket with a multiscale-multiphysics framework

The Water-Cooled Lithium-Lead (WCLL) blanket concept is one of four possible embodiments under consideration for the European Demonstration Power Plant DEMO, beyond ITER. It is based on a water-cooled first wall and blanket structure with the RAFMS (Reduced Activation Ferritic/Martensitic Steel, Eurofer 97) as structural material and PbLi as neutron multiplier and tritium breeder. We first develop a multiphysics modeling framework in order to optimize the design and achieve long blanket lifetime, maintainability, and high reliability. The approach is demonstrated by coupling Computational Fluid Dynamics (CFD), heat transfer and solid mechanics. The multiphysics framework is then utilized to provide the loading conditions for a multiscale design approach particularly focused on modeling the effects of both steady-state and operational transients on the structural integrity of the First Wall (FW). The methodology builds on the ITER Structural Design Criteria for In-Vessel Components (ISDC-IC), then extending the criteria to progressively incorporate continuum plasticity models and microstructure-based representation of deformation and fracture. We finally introduce advanced fracture mechanics concepts based on a Materials-Specific Failure Assessment Diagram (MS-FAD), in which precise calculations of the J-integral of an elasto-plastic material is required. This is especially important for the water-cooled design because of the limited fracture toughness of Eurofer 97 below 300 °C. The methodology has been applied to screen the FW/Blanket performance of a range of steady state heat flux assumed to be applied during the plasma pulse flat top, and up to 0.8 MW/m 2 . It has become clear that the radiation steady state load will be lower than the values screened here, while the local heat flux due to particle loads will lead to global higher values than the values investigated here. The methodology will be further fine-tuned to specific loading conditions in order to assess actual performance limits of the DEMO WCLL BB.

European DEMO↗

Unique W-Shape Y6 isomer as effective solid additive for High-Performance PM6:Y6 polymer solar cells

The current top-performing polymer solar cell (PSC) systems are mostly based on PM6:Y6 host blend. To date, numerous materials have been explored as the third component for these systems to form ternary blends or as additives. Vitrification agents are a group of additives proved to be useful in affecting the morphology of organic semiconductors. Here to design a suitable vitrification agent for non-fullerene electron acceptor Y6, an isomer strategy was explored where the thienothiophene wings of the Y6 molecule were inversed to form a W-shaped Y6 isomer of i-Y6. It was found that i-Y6 crystallized poorly with a different packing style than Y6 and could blend well in amorphous phase of Y6. These properties enabled i-Y6 to finely tune the morphology of PM6:Y6 blend at low additive dosages. The power conversion efficiency (PCE) of PM6:Y6 based PSC was raised from 16.827% to 17.433% at a dosage of Y6:i-Y6 ratio of 24:1. This work demonstrated isomerization as a viable strategy for developing solid additives for high performance PSC blends and vitrification agents as effective additives for tuning morphology and improving performances of PSCs.

36 MATERIALS SCIENCE↗

Limits of multifunctionality in tunable networks

Nature is rife with networks that are functionally optimized to propagate inputs to perform specific tasks. Whether via genetic evolution or dynamic adaptation, many networks create functionality by locally tuning interactions between nodes. Here we explore this behavior in two contexts: strain propagation in mechanical networks and pressure redistribution in flow networks. By adding and removing links, we are able to optimize both types of networks to perform specific functions. We define a single function as a tuned response of a single “target” link when another, predetermined part of the network is activated. Using network structures generated via such optimization, we investigate how many simultaneous functions such networks can be programed to fulfill. We find that both flow and mechanical networks display qualitatively similar phase transitions in the number of targets that can be tuned, along with the same robust finite-size scaling behavior. We discuss how these properties can be understood in the context of constraint–satisfaction problems.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗

Transformers and Long Short-Term Memory Transfer Learning for GenIV Reactor Temperature Time Series Forecasting

Automated monitoring of the coolant temperature can enable autonomous operation of generation IV reactors (GenIV), thus reducing their operating and maintenance costs. Automation can be accomplished with machine learning (ML) models trained on historical sensor data. However, the performance of ML usually depends on the availability of large amount of training data, which is difficult to obtain for GenIV, as this technology is still under development. We propose the use of transfer learning (TL), which involves utilizing knowledge across different domains, to compensate for this lack of training data. TL can be used to create pre-trained ML models with data from small-scale research facilities, which can then be fine-tuned to monitor GenIV reactors. In this work, we develop pre-trained Transformer and long short-term memory (LSTM) networks by training them on temperature measurements from thermal hydraulic flow loops operating with water and Galinstan fluids at room temperature at Argonne National Laboratory. The pre-trained models are then fine-tuned and re-trained with minimal additional data to perform predictions of the time series of high temperature measurements obtained from the Engineering Test Unit (ETU) at Kairos Power. The performance of the LSTM and Transformer networks is investigated by varying the size of the lookback window and forecast horizon. The results of this study show that LSTM networks have lower prediction errors than Transformers, but LSTM errors increase more rapidly with increasing lookback window size and forecast horizon compared to the Transformer errors.

LSTM↗

Optimizing Irregular Applications for Energy and Performance on the Tilera Many-core Architecture

Optimizing applications simultaneously for energy and performance is a complex problem. High performance, parallel, irregular applications are notoriously hard to optimize due to their data-dependent memory accesses, lack of structured locality and complex data structures and code patterns. Irregular kernels are growing in importance in applications such as machine learning, graph analytics and combinatorial scientific computing. Performance- and energy-efficient implementation of these kernels on modern, energy efficient, multicore and many-core platforms is therefore an important and challenging problem. We present results from optimizing two irregular applications { the Louvain method for community detection (Grappolo), and high-performance conjugate gradient (HPCCG) { on the Tilera many-core system. We have significantly extended MIT's OpenTuner auto-tuning framework to conduct a detailed study of platform-independent and platform-specific optimizations to improve performance as well as reduce total energy consumption. We explore the optimization design space along three dimensions: memory layout schemes, compiler-based code transformations, and optimization of parallel loop schedules. Using auto-tuning, we demonstrate whole node energy savings of up to 41% relative to a baseline instantiation, and up to 31% relative to manually optimized variants.

irregular applications, energy opttimization, many↗

Surrogate-Based Autotuning for Randomized Sketching Algorithms in Regression Problems

Algorithms from Randomized Numerical Linear Algebra (RandNLA) are known to be effective in handling high-dimensional computational problems, providing high-quality empirical performance as well as strong probabilistic guarantees. However, their practical application is complicated by the fact that the user needs to set various algorithm-specific tuning parameters which are different from those used in traditional NLA. This paper demonstrates how a surrogate-based autotuning approach can be used to address fundamental problems of parameter selection in RandNLA algorithms. In particular, we provide a detailed investigation of surrogate-based autotuning for sketch-and-precondition (SAP)-based randomized least squares methods, which have been one of the great success stories in modern RandNLA. Empirical results show that our surrogate-based autotuning approach can achieve near-optimal performance with much less tuning cost than a random search (up to about 7.6x fewer trials of different parameter configurations). Moreover, while our experiments focus on least squares, our results demonstrate a general-purpose autotuning pipeline applicable to any kind of RandNLA algorithm.

Cho, Younghyun↗