Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Handling”

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 1,711 records · Page 95
All sourcesNASA NTRSDOE OSTI
DOE OSTI2013

Surface characterization of nanomaterials and nanoparticles: Important needs and challenging opportunities

This review examines characterization challenges inherently associated with understanding nanomaterials and the roles surface and interface characterization methods can play in meeting some of the challenges. In parts of the research community, there is growing recognition that studies and published reports on the properties and behaviors of nanomaterials often have reported inadequate or incomplete characterization. As a consequence, the true value of the data in these reports is, at best, uncertain. With the increasing importance of nanomaterials in fundamental research and technological applications, it is desirable that researchers from the wide variety of disciplines involved recognize the nature of these often unexpected challenges associated with reproducible synthesis and characterization of nanomaterials, including the difficulties of maintaining desired materials properties during handling and processing due to their dynamic nature. It is equally valuable for researchers to understand how characterization approaches (surface and otherwise) can help to minimize synthesis surprises and to determine how (and how quickly) materials and properties change in different environments. Appropriate application of traditional surface sensitive analysis methods (including x-ray photoelectron and Auger electron spectroscopies, scanning probe microscopy, and secondary ion mass spectroscopy) can provide information that helps address several of the analysis needs. In many circumstances, extensions of traditional data analysis can provide considerably more information than normally obtained from the data collected. Less common or evolving methods with surface selectivity (e.g., some variations of nuclear magnetic resonance, sum frequency generation, and low and medium energy ion scattering) can provide information about surfaces or interfaces in working environments (operando or in situ) or information not provided by more traditional methods. Although these methods may require instrumentation or expertise not generally available, they can be particularly useful in addressing specific questions, and examples of their use in nanomaterial research are presented.

77 NANOSCIENCE AND NANOTECHNOLOGY↗
DOE OSTI2016

Provenance information as a tool for addressing engineered nanoparticle reproducibility challenges

Nanoparticles of various types are of increasing research and technological importance in biological and other applications. Difficulties in the production and delivery of nanoparticles with consistent and well defined properties appear in many forms and have a variety of causes. Among several issues are those associated with incomplete information about the history of particles involved in research studies including the synthesis method, sample history after synthesis including time and nature of storage and the detailed nature of any sample processing or modification. In addition, the tendency of particles to change with time or environmental condition suggests that the time between analysis and application is important and some type of consistency or verification process can be important. The essential history of a set of particles can be identified as provenance information tells the origin or source of a batch of nano-objects along with information related to handling and any changes that may have taken place since it was originated. A record of sample provenance information for a set of particles can play a useful role in identifying some of the sources and decreasing the extent of particle variability and the observed lack of reproducibility observed by many researchers.

Environmental Molecular Sciences Laboratory↗
DOE OSTI2020

Advanced modeling of field enhanced thermionic emission

Shaped emitters are of interest to a broad range of applications in vacuum electronic devices. In particular, thermionic energy converters (TECs) take advantage of shaped emitters to increase the local surface field, thereby extracting more current for a given cathode temperature and applied voltage. However, modeling these devices is challenging; Warp [J.-L. Vay, D. P. Grote, R. H. Cohen, and A. Friedman, Comput. Sci. Discov. 5, 014019 (2012)] is a fully 3D particle-in-cell code capable of handling a wide range of physics problems and is well suited to modeling TECs. Additionally, recent improvements to Warp have enabled the accurate modeling of emitters with arbitrary curved surfaces. Specifically, the inclusion of subgrid resolution for computing the electrostatic potential and the ability to apply mesh refinement for specific areas of interest allow for a more accurate solution to the fields on these surfaces. These improvements coupled with Warp’s ability to handle variable particle weights make it an ideal candidate for simulating these complex devices. In this paper, the authors study the applicability of different subgrid configurations for simulating shaped emission surfaces and field convergence for different mesh-refinement techniques. They then implement a custom weighting algorithm that allows for uniform sampling of emission surfaces with a large variation in the surface electric field. They then use this algorithm to study emission for curved emitters in both the field-enhancement regime and the space-charge regime.

Edelen, Jonathan P. (ORCID:0000000215180652)↗
DOE OSTI2014

Analytical, experimental, and Monte Carlo system response matrix for pinhole SPECT reconstruction

Purpose: To assess the performance of two approaches to the system response matrix (SRM) calculation in pinhole single photon emission computed tomography (SPECT) reconstruction. Methods: Evaluation was performed using experimental data from a low magnification pinhole SPECT system that consisted of a rotating flat detector with a monolithic scintillator crystal. The SRM was computed following two approaches, which were based on Monte Carlo simulations (MC-SRM) and analytical techniques in combination with an experimental characterization (AE-SRM). The spatial response of the system, obtained by using the two approaches, was compared with experimental data. The effect of the MC-SRM and AE-SRM approaches on the reconstructed image was assessed in terms of image contrast, signal-to-noise ratio, image quality, and spatial resolution. To this end, acquisitions were carried out using a hot cylinder phantom (consisting of five fillable rods with diameters of 5, 4, 3, 2, and 1 mm and a uniform cylindrical chamber) and a custom-made Derenzo phantom, with center-to-center distances between adjacent rods of 1.5, 2.0, and 3.0 mm. Results: Good agreement was found for the spatial response of the system between measured data and results derived from MC-SRM and AE-SRM. Only minor differences for point sources at distances smaller than the radius of rotation and large incidence angles were found. Assessment of the effect on the reconstructed image showed a similar contrast for both approaches, with values higher than 0.9 for rod diameters greater than 1 mm and higher than 0.8 for rod diameter of 1 mm. The comparison in terms of image quality showed that all rods in the different sections of a custom-made Derenzo phantom could be distinguished. The spatial resolution (FWHM) was 0.7 mm at iteration 100 using both approaches. The SNR was lower for reconstructed images using MC-SRM than for those reconstructed using AE-SRM, indicating that AE-SRM deals better with the projection noise than MC-SRM. Conclusions: The authors' findings show that both approaches provide good solutions to the problem of calculating the SRM in pinhole SPECT reconstruction. The AE-SRM was faster to create and handle the projection noise better than MC-SRM. Nevertheless, the AE-SRM required a tedious experimental characterization of the intrinsic detector response. Creation of the MC-SRM required longer computation time and handled the projection noise worse than the AE-SRM. Nevertheless, the MC-SRM inherently incorporates extensive modeling of the system and therefore experimental characterization was not required.

62 RADIOLOGY AND NUCLEAR MEDICINE↗
DOE OSTI2014

SU-E-T-395: Multi-GPU-Based VMAT Treatment Plan Optimization Using a Column-Generation Approach

Purpose: GPU has been employed to speed up VMAT optimizations from hours to minutes. However, its limited memory capacity makes it difficult to handle cases with a huge dose-deposition-coefficient (DDC) matrix, e.g. those with a large target size, multiple arcs, small beam angle intervals and/or small beamlet size. We propose multi-GPU-based VMAT optimization to solve this memory issue to make GPU-based VMAT more practical for clinical use. Methods: Our column-generation-based method generates apertures sequentially by iteratively searching for an optimal feasible aperture (referred as pricing problem, PP) and optimizing aperture intensities (referred as master problem, MP). The PP requires access to the large DDC matrix, which is implemented on a multi-GPU system. Each GPU stores a DDC sub-matrix corresponding to one fraction of beam angles and is only responsible for calculation related to those angles. Broadcast and parallel reduction schemes are adopted for inter-GPU data transfer. MP is a relatively small-scale problem and is implemented on one GPU. One headand- neck cancer case was used for test. Three different strategies for VMAT optimization on single GPU were also implemented for comparison: (S1) truncating DDC matrix to ignore its small value entries for optimization; (S2) transferring DDC matrix part by part to GPU during optimizations whenever needed; (S3) moving DDC matrix related calculation onto CPU. Results: Our multi-GPU-based implementation reaches a good plan within 1 minute. Although S1 was 10 seconds faster than our method, the obtained plan quality is worse. Both S2 and S3 handle the full DDC matrix and hence yield the same plan as in our method. However, the computation time is longer, namely 4 minutes and 30 minutes, respectively. Conclusion: Our multi-GPU-based VMAT optimization can effectively solve the limited memory issue with good plan quality and high efficiency, making GPUbased ultra-fast VMAT planning practical for real clinical use.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2014

SU-E-T-590: An Activation Study of Materials and Devices Present in a Proton Treatment Room

Purpose: The use of protons for radiation therapy is growing rapidly. One consequence of protons interacting with different media is activation. These nuclear reactions induced by the protons, scattered neutrons, and gamma rays, activate different materials encountered, particularly by the therapists. The purpose of this study was to examine the derived nuclides from the activation, and also the decay rate. Methods: The study was conducted in our proton therapy facility. Protons are derived from a synchrocyclotron and pass through field shipping systems, apertures, and range compensators to define the beam within the patient.Included materials of concerns measured; the patient support couch, sheet rock in the wall, solid plastics used for quality assurance and dosimetry, and the passive scattering system itself, which includes brass apertures, and Lucite or blue wax compensators. All devices were studied post irradiation using gamma spectroscopy to determine the nuclides, and a sodium iodine scintillation detector to measure decay, particularly when the dose rate fell to background levels. Results: We have also determined from the measurements we will maintain brass apertures for three months before sending them for scrap. Some of the radionuclides arrived from these measurements included Na-22 for the blue wax compensator, C1-34m for the sheetrock, and Sc-44 and Co-60 for the brass apertures. We found compensators made out of Lucite or wax decayed to background in 2 hours. The patient support couch decayed to background in approximately 40 minutes, and sheet rock decayed in 80 minutes. In terms of the aperture layers, the most proximal aperture slab had much higher activity than the distal slab. Also the proximal sides of the slabs were much more activate than the distal. Conclusion: We have given proper instruction to therapists performing quality assurance in terms of the handled plastics, and to handle apertures rapidly as possible.

07 ISOTOPE AND RADIATION SOURCES↗
DOE OSTI2015

Multi-GPU implementation of a VMAT treatment plan optimization algorithm

Purpose: Volumetric modulated arc therapy (VMAT) optimization is a computationally challenging problem due to its large data size, high degrees of freedom, and many hardware constraints. High-performance graphics processing units (GPUs) have been used to speed up the computations. However, GPU’s relatively small memory size cannot handle cases with a large dose-deposition coefficient (DDC) matrix in cases of, e.g., those with a large target size, multiple targets, multiple arcs, and/or small beamlet size. The main purpose of this paper is to report an implementation of a column-generation-based VMAT algorithm, previously developed in the authors’ group, on a multi-GPU platform to solve the memory limitation problem. While the column-generation-based VMAT algorithm has been previously developed, the GPU implementation details have not been reported. Hence, another purpose is to present detailed techniques employed for GPU implementation. The authors also would like to utilize this particular problem as an example problem to study the feasibility of using a multi-GPU platform to solve large-scale problems in medical physics. Methods: The column-generation approach generates VMAT apertures sequentially by solving a pricing problem (PP) and a master problem (MP) iteratively. In the authors’ method, the sparse DDC matrix is first stored on a CPU in coordinate list format (COO). On the GPU side, this matrix is split into four submatrices according to beam angles, which are stored on four GPUs in compressed sparse row format. Computation of beamlet price, the first step in PP, is accomplished using multi-GPUs. A fast inter-GPU data transfer scheme is accomplished using peer-to-peer access. The remaining steps of PP and MP problems are implemented on CPU or a single GPU due to their modest problem scale and computational loads. Barzilai and Borwein algorithm with a subspace step scheme is adopted here to solve the MP problem. A head and neck (H and N) cancer case is then used to validate the authors’ method. The authors also compare their multi-GPU implementation with three different single GPU implementation strategies, i.e., truncating DDC matrix (S1), repeatedly transferring DDC matrix between CPU and GPU (S2), and porting computations involving DDC matrix to CPU (S3), in terms of both plan quality and computational efficiency. Two more H and N patient cases and three prostate cases are used to demonstrate the advantages of the authors’ method. Results: The authors’ multi-GPU implementation can finish the optimization process within ∼1 min for the H and N patient case. S1 leads to an inferior plan quality although its total time was 10 s shorter than the multi-GPU implementation due to the reduced matrix size. S2 and S3 yield the same plan quality as the multi-GPU implementation but take ∼4 and ∼6 min, respectively. High computational efficiency was consistently achieved for the other five patient cases tested, with VMAT plans of clinically acceptable quality obtained within 23–46 s. Conversely, to obtain clinically comparable or acceptable plans for all six of these VMAT cases that the authors have tested in this paper, the optimization time needed in a commercial TPS system on CPU was found to be in an order of several minutes. Conclusions: The results demonstrate that the multi-GPU implementation of the authors’ column-generation-based VMAT optimization can handle the large-scale VMAT optimization problem efficiently without sacrificing plan quality. The authors’ study may serve as an example to shed some light on other large-scale medical physics problems that require multi-GPU techniques.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2016

The rotate-plus-shift C-arm trajectory. Part I. Complete data with less than 180° rotation

Purpose: In the last decade, C-arm-based cone-beam CT became a widely used modality for intraoperative imaging. Typically a C-arm CT scan is performed using a circular or elliptical trajectory around a region of interest. Therefore, an angular range of at least 180° plus fan angle must be covered to ensure a completely sampled data set. However, mobile C-arms designed with a focus on classical 2D applications like fluoroscopy may be limited to a mechanical rotation range of less than 180° to improve handling and usability. The method proposed in this paper allows for the acquisition of a fully sampled data set with a system limited to a mechanical rotation range of at least 180° minus fan angle using a new trajectory design. This enables CT like 3D imaging with a wide range of C-arm devices which are mainly designed for 2D imaging. Methods: The proposed trajectory extends the mechanical rotation range of the C-arm system with two additional linear shifts. Due to the divergent character of the fan-beam geometry, these two shifts lead to an additional angular range of half of the fan angle. Combining one shift at the beginning of the scan followed by a rotation and a second shift, the resulting rotate-plus-shift trajectory enables the acquisition of a completely sampled data set using only 180° minus fan angle of rotation. The shifts can be performed using, e.g., the two orthogonal positioning axes of a fully motorized C-arm system. The trajectory was evaluated in phantom and cadaver examinations using two prototype C-arm systems. Results: The proposed trajectory leads to reconstructions without limited angle artifacts. Compared to the limited angle reconstructions of 180° minus fan angle, image quality increased dramatically. Details in the rotate-plus-shift reconstructions were clearly depicted, whereas they are dominated by artifacts in the limited angle scan. Conclusions: The method proposed here employs 3D imaging using C-arms with less than 180° rotation range adding full 3D functionality to a C-arm device retaining both handling comfort and the usability of 2D imaging. This method has a clear potential for clinical use especially to meet the increasing demand for an intraoperative 3D imaging.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2016

SU-F-R-18: Updates to the Computational Environment for Radiological Research for Image Analysis

Purpose: To present new tools in CERR for Texture Analysis and Visualization. Method: (1) Quantitative Image Analysis: We added the ability to compute Haralick texture features based on local neighbourhood. The Texture features depend on many parameters used in their derivation. For example: (a) directionality, (b) quantization of image, (c) patch-size for the neighborhood, (d) handling of the edge voxels within the region of interest, (e) Averaging co-occurance matrix vs texture features for different directions etc. A graphical user interface was built to set these parameters and then visualize their impact on the resulting texture maps. The entire functionality was written in Matlab. Array indexing was used to speed up the texture calculation. The computation speed is very competitive with the ITK library. Moreover, our implementation works with multiple CPUs and the computation time can be further reduced by using multiple processor threads. In order to reduce the Haralick texture maps into scalar features, we propose the use of Texture Volume Histograms. This lets users make use of the entire distribution of texture values within the region of interest rather than using just the mean and the standard deviations. (2) Qualitative/Visualization tools: The derived texture maps are stored as a new scan (derived) within CERR’s planC data structure. A display that compares various scans was built to show the raw image and the derived texture maps side-by-side. These images are positionally linked and can be navigated together. CERR’s graphics handling was updated and sped-up to be compatible with the newer Matlab versions. As a result, the users can use (a) different window levels and colormaps for different viewports, (b) click-and-drag or use mouse scroll-wheel to navigate slices. Results: The new features and updates are available via https://www.github.com/adityaapte/cerr . Conclusion: Features added to CERR increase its utility in Radiomics and Outcomes modeling.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2016

A pencil beam dose calculation model for CyberKnife system

Purpose: CyberKnife system is initially equipped with fixed circular cones for stereotactic radiosurgery. Two dose calculation algorithms, Ray-Tracing and Monte Carlo, are available in the supplied treatment planning system. A multileaf collimator system was recently introduced in the latest generation of system, capable of arbitrarily shaped treatment field. The purpose of this study is to develop a model based dose calculation algorithm to better handle the lateral scatter in an irregularly shaped small field for the CyberKnife system. Methods: A pencil beam dose calculation algorithm widely used in linac based treatment planning system was modified. The kernel parameters and intensity profile were systematically determined by fitting to the commissioning data. The model was tuned using only a subset of measured data (4 out of 12 cones) and applied to all fixed circular cones for evaluation. The root mean square (RMS) of the difference between the measured and calculated tissue-phantom-ratios (TPRs) and off-center-ratio (OCR) was compared. Three cone size correction techniques were developed to better fit the OCRs at the penumbra region, which are further evaluated by the output factors (OFs). The pencil beam model was further validated against measurement data on the variable dodecagon-shaped Iris collimators and a half-beam blocked field. Comparison with Ray-Tracing and Monte Carlo methods was also performed on a lung SBRT case. Results: The RMS between the measured and calculated TPRs is 0.7% averaged for all cones, with the descending region at 0.5%. The RMSs of OCR at infield and outfield regions are both at 0.5%. The distance to agreement (DTA) at the OCR penumbra region is 0.2 mm. All three cone size correction models achieve the same improvement in OCR agreement, with the effective source shift model (SSM) preferred, due to their ability to predict more accurately the OF variations with the source to axis distance (SAD). In noncircular field validation, the pencil beam calculated results agreed well with the film measurement of both Iris collimators and the half-beam blocked field, fared much better than the Ray-Tracing calculation. Conclusions: The authors have developed a pencil beam dose calculation model for the CyberKnife system. The dose calculation accuracy is better than the standard linac based system because the model parameters were specifically tuned to the CyberKnife system and geometry correction factors. The model handles better the lateral scatter and has the potential to be used for the irregularly shaped fields. Comprehensive validations on MLC equipped system are necessary for its clinical implementation. It is reasonably fast enough to be used during plan optimization.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2025

A Linear-Complexity Tensor Butterfly Algorithm for Compressing High-Dimensional Oscillatory Integral Operators

This paper presents a multilevel tensor compression algorithm called tensor butterfly algorithm for efficiently representing large-scale and high-dimensional oscillatory integral operators, including Green's functions for wave equations and integral transforms such as Radon transforms and Fourier transforms. The proposed algorithm leverages a tensor extension of the so-called complementary low-rank property of existing matrix butterfly algorithms. The algorithm partitions the discretized integral operator tensor into subtensors of multiple levels and factorizes each subtensor at the middle level as a Tucker-type interpolative decomposition, whose factor matrices are formed in a multilevel fashion. For a d-dimensional (d > 1) integral operator discretized into a 2d-mode tensor with n2d entries, the overall CPU time and memory requirement scale as O(nd), in stark contrast to the O(nd log n) complexity of existing matrix algorithms such as matrix butterfly algorithms and fast Fourier transforms (FFTs), where n is the number of points per direction. When comparing with other tensor algorithms such as quantized tensor train (QTT), the proposed algorithm also shows superior CPU and memory performance for tensor contraction. Remarkably, the tensor butterfly algorithm can efficiently model high-frequency Green's function interactions between two unit cubes, each spanning 512 wavelengths per direction, which represents problems of scale over 512× larger than that existing butterfly algorithms can handle, with the same amount of computation resources. On the other hand, for a problem representing 64 wavelengths per direction, which is the largest size existing algebraic matrix algorithms can handle, our tensor butterfly algorithm exhibits 200x speedups and 30× memory reduction compared with existing ones. Moreover, the tensor butterfly algorithm also permits O(nd)-complexity FFTs and Radon transforms up to d = 6 dimensions.

Kielstra, P Michael↗
DOE OSTI2023

Benchmarking for AI for Science

AI has been instrumental for recent developments in a number of domains of the sciences. With several hundred machine learning (ML) algorithms and models, and numerous AI-specific hardware platforms, a common quest for all scientists working on AI for Science is around the selection of machine learning algorithm(s) to solve their domain-specific scientific problems. A number of different initiatives around AI Benchmarking have been set up and have been useful in understanding the benefits of different ML algorithms for different tasks.However, with the majority of these AI Benchmarking initiatives focusing on the conventional notions of benchmarking, where the focus is purely runtime performance (such as training time or inference time), their suitability for benchmarking different ML algorithms for solving scientific problems has been viewed as a performance problem even though both are hardly the same. To make reasonable, explainable, and justifiable advancements in science using AI, it is critical to focus on the merits of these algorithms in handling different domain science problems. In other words, more emphasis must be given on Benchmarking for AI for Science than AI Benchmarking. The vision of the former is not only to assess the performance of ML algorithms, but also to assess, and understand the benefits and merits of different ML algorithms in handling scientific problems. Benchmarking for AI for Science, instead of pure performance focused AI Benchmarking, has several benefits: (i) it has the potential to offer advances in the sciences, much more rapidly than through pure performance-based AI methods, (ii) it will encourage the community to focus on developing better domain-specific AI techniques, particularly given the provision for being able to benchmark different techniques, and (iii) it will encourage hardware manufacturers to focus on developing science-specific hardware subsystems.

Thiyagalingam, Jeyan↗
DOE OSTI2017

A Pattern Language for High-Performance Computing Resilience

High-performance computing systems (HPC) provide powerful capabilities for modeling, simulation, and data analytics for a broad class of computational problems. They enable extreme performance of the order of quadrillion floating-point arithmetic calculations per second by aggregating the power of millions of compute, memory, networking and storage components. With the rapidly growing scale and complexity of HPC systems for achieving even greater performance, ensuring their reliable operation in the face of system degradations and failures is a critical challenge. System fault events often lead the scientific applications to produce incorrect results, or may even cause their untimely termination. The sheer number of components in modern extreme-scale HPC systems and the complex interactions and dependencies among the hardware and software components, the applications, and the physical environment makes the design of practical solutions that support fault resilience a complex undertaking. To manage this complexity, we developed a methodology for designing HPC resilience solutions using design patterns. We codified the well-known techniques for handling faults, errors and failures that have been devised, applied and improved upon over the past three decades in the form of design patterns. In this paper, we present a pattern language to enable a structured approach to the development of HPC resilience solutions. The pattern language reveals the relations among the resilience patterns and provides the means to explore alternative techniques for handling a specific fault model that may have different efficiency and complexity characteristics. Using the pattern language enables the design and implementation of comprehensive resilience solutions as a set of interconnected resilience patterns that can be instantiated across layers of the system stack.

Hukerikar, Saurabh↗
DOE OSTI2018

Pattern-based Modeling of Multiresilience Solutions for High-Performance Computing

Resiliency is the ability of large-scale high-performance computing (HPC) applications to gracefully handle errors, and recover from failures. In this paper, we propose a pattern-based approach to constructing resilience solutions that handle multiple error modes. Using resilience patterns, we evaluate the performance and reliability characteristics of detection, containment and mitigation techniques for transient errors that cause silent data corruptions and techniques for fail-stop errors that result in process failures. We demonstrate the design and implementation of the multiresilience solution based on patterns instantiated across multiple layers of the system stack. The patterns are integrated to work together to achieve resiliency to different error types in a performance-efficient manner.

Ashraf, Rizwan↗
DOE OSTI2018

Comparative Analysis of Soft-Error Detection Strategies: A Case Study with Iterative Methods

Undetected soft errors caused by transient bit flips can lead to silent data corruption (SDC), an undesirable outcome where invalid results pass for valid ones. This has motivated the design of soft error detectors to minimize SDCs. However, the detectors have been studied under different contexts, making comparative evaluation difficult. In this paper, we present the first comprehensive evaluation of four online soft error detection techniques in detecting the adverse impact of soft errors on iterative methods. We observe that, across five iterative methods, the detectors studied achieve high but not perfect detection rates. To understand the potential for improved detection, we evaluate a machine-learning based detector that takes as features that are the runtime features observed by the individual detectors to arrive at their conclusions. Our evaluation demonstrates improved but still far from perfect detection accuracy for the machine learning based detectors. This extensive evaluation demonstrates the need for designing error detectors to handle the evolutionary behavior exhibited by iterative solvers.

Silent data corruption, Soft error detection, Iter↗
DOE OSTI2021

Can Applications Recover from fsync Failures?

We analyze how file systems and modern data-intensive applications react to fsync failures. First, we characterize how three Linux file systems (ext4, XFS, Btrfs) behave in the presence of failures. We find commonalities across file systems (pages are always marked clean, certain block writes always lead to unavailability) as well as differences (page content and failure reporting is varied). Next, we study how five widely used applications (PostgreSQL, LMDB, LevelDB, SQLite, Redis) handle fsync failures. Our findings show that although applications use many failure-handling strategies, none are sufficient: fsync failures can cause catastrophic outcomes such as data loss and corruption. Our findings have strong implications for the design of file systems and applications that intend to provide strong durability guarantees.

Computer Science↗
DOE OSTI2021

cuTS: Scaling Subgraph Isomorphism on Distributed Multi-GPUSystems Using Trie Based Data Structure

Subgraph isomorphism is a pattern-matching algorithm widely used in many domains such as chem-informatics, bioinformatics, databases, and social network analysis. It is computationally expensive and is a proven NP-hard problem. The massive parallelism offered by the GPU hardware is well suited for solving the subgraph isomorphism. However, current GPU implementations are far from the achievable performance. Moreover, the enormous memory requirement of current approaches limits the problem size that can be handled. This work analyzes the fundamental challenges associated with processing the subgraph isomorphism on GPUs and develops an efficient GPU hardware-aware implementation. We also develop a new GPU-friendly trie-based data structure to drastically reduce the intermediate storage space requirement. Hence, our approach runs larger benchmarks than the competitors. We also develop the first distributed sub-graph isomorphism algorithm for GPUs. Our experimental evaluation section demonstrates the efficacy of our approach by comparing the execution time and number of cases that we can handle against the state-of-the-art GPU implementations.

Xiang, Lizhi↗
DOE OSTI2022

Attention-based Aspect Reasoning for Knowledge Base Question Answering on Clinical Notes

Question Answering (QA) in clinical notes has gained a lot of attention in the past few years. Existing machine reading comprehension approaches in clinical domain can only handle questions about a single block of clinical texts and fail to retrieve information about different patients and clinical notes. To handle more complex questions, we aim at creating knowledge base from clinical notes to link different patients and clinical notes, and performing knowledge base question answering (KBQA). Based on the expert annotations in n2c2, we first created the ClinicalKBQA dataset that includes 8,952 QA pairs and covers questions about seven medical topics through 322 question templates. Then, we proposed an attention-based aspect reasoning (AAR) method for KBQA and investigated the impact of different aspects of answers (e.g., entity, type, path, and context) for prediction. The AAR method achieves better performance due to the well-designed encoder and attention mechanism. In the experiments, we find that both aspects, type and path, enable the model to identify answers satisfying the general conditions and produce lower precision and higher recall. On the other hand, the aspects, entity and context, limit the answers by node-specific information and lead to higher precision and lower recall.

Wang, Ping↗