Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Kernel”

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 991 records · Page 55
All sourcesNASA NTRSDOE OSTI
DOE OSTI2015

SU-E-T-236: Deconvolution of the Total Nuclear Cross-Sections of Therapeutic Protons and the Characterization of the Reaction Channels

Purpose: The knowledge of the total nuclear cross-section Qtot(E) of therapeutic protons Qtot(E) provides important information in advanced radiotherapy with protons, such as the decrease of fluence of primary protons, the release of secondary particles (neutrons, protons, deuterons, etc.), and the production of nuclear fragments (heavy recoils), which usually undergo β+/− decay by emission of γ-quanta. Therefore determination of Qtot(E) is an important tool for sophisticated calculation algorithms of dose distributions. This cross-section can be determined by a linear combination of shifted Gaussian kernels and an error-function. The resonances resulting from deconvolutions in the energy space can be associated with typical nuclear reactions. Methods: The described method of the determination of Qtot(E) results from an extension of the Breit-Wigner formula and a rather extended version of the nuclear shell theory to include nuclear correlation effects, clusters and highly excited/virtually excited nuclear states. The elastic energy transfer of protons to nucleons (the quantum numbers of the target nucleus remain constant) can be removed by the mentioned deconvolution. Results: The deconvolution of the term related to the error-function of the type cerf*er((E-ETh)/σerf] is the main contribution to obtain various nuclear reactions as resonances, since the elastic part of energy transfer is removed. The nuclear products of various elements of therapeutic interest like oxygen, calcium are classified and calculated. Conclusions: The release of neutrons is completely underrated, in particular, for low-energy protons. The transport of seconary particles, e.g. cluster formation by deuterium, tritium and α-particles, show an essential contribution to secondary particles, and the heavy recoils, which create γ-quanta by decay reactions, lead to broadening of the scatter profiles. These contributions cannot be accounted for by one single Gaussian kernel for the description of lateral scatter.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2015

SU-F-BRD-06: Robust Dose Calculation in Intensity Modulated Proton Therapy

Purpose: Commissioning data for intensity modulated proton therapy (IMPT) must be post-processed by fits to ad-hoc functions to derive the dose calculation kernel parameters in a treatment planning system (TPS). Whether from experimental measurement or Monte Carlo simulation, the limited and noisy nature of such data makes this task very challenging. We present a method to improve the modeling of the lateral dose distribution of clinical energy proton beams in water to commission an in-house IMPT dose calculation engine. Methods: A linear sum of three Gaussian distribution functions was fitted to the lateral dose data in logarithmic scale. Starting values of fitting solutions were determined from the Generalized Highland Approximation. We exhaustively optimized the combinations of data weights with upper bounds of the fitting solutions to minimize confidence intervals of the fitting solutions while maintaining the coefficient of determination (R{sup 2}). Results: Across all energies, average confidence bounds improved 72.88% [Max: 88.28%, Min: 55.05%] for small angle coulomb scattering, 114.25% [409.13%, 66.72%,] for nuclear scattering, and 68.66% [141.09%, 33.27%] for large angle coulomb scattering, while the coefficients of determination of the fits (R{sup 2}) remained comparable. On average R {sup 2} only changed 0.18% and were very close to 1 (approx. 0.999). Wilcoxon signed rank tests comparing unweighted/unbounded fits with weighted/bounded fits averaged 0.0146 (Max: 0.177, Min: 7.05×10−{sup 7}) for small angle Coulomb, 0.0903 (0.945, 7.05×10−{sup 7}) for nuclear, and 0.254 (0.871, 1.86×10−{sup 6}) for large angle Coulomb scattering. This allows rejection of the null hypothesis for small angle Coulomb scattering at the 0.015 level and nuclear interaction at the 0.1 level. Conclusion: Optimal weights assigned to IMPT lateral dose data minimized fitting to stochastic noise in the tail region. Optimizing the upper bounds of fitting parameters improved the robustness of fitting solutions. Both led to a more physically meaningful dose calculation kernel for the TPS. NIH/NCI K25CA168984, Eagles Cancer Research Career Development, The Lawrence W. and Marilyn W. Matteson Fund for Cancer Research, Mayo ASU Seed Grant, and The Kemper Marley Foundation.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2015

SU-F-BRA-15: Physical Aspects and Clinical Applications of Post-Treatment Yttrium-90 PET-Based Dosimetry

Purpose: The accuracy of PET-based post-treatment dosimetry of yttrium-90 microspheres has been improving over the past decade and is now at a stage, permitting volumetric dose-outcome studies. We outline the recent advances and identify the physical limitations to the accuracy of the dose calculations. Methods: Convolution of the measured PET activity density distribution with a pre-calculated voxel-dose-kernel (VDK) is the most widely used method for dose reconstruction. Therefore, accurate knowledge of the beta+ branching ratio as well as the micro-dosimetric characteristics of electron interactions within the microsphere is essential for the computation of the dose kernel. We implement a model of the microspheres used in our clinic to calculate the modified electron energy spectrum at the microsphere’s surface and to determine the impact of self-shielding on the dose reconstruction. The three-dimensional dose distributions obtained for 10 patients treated with radio-embolization with yttrium-90 microspheres are evaluated and various DVH markers are investigated for correlation with outcome. Results: The methods for reducing the overall computation uncertainty are systematically outlined in this presentation. Since the latest experimental data on yttrium-90 beta+ branching ratio has a relative uncertainty of 1.5%, all contributing factors derived from Monte Carlo simulations must be brought to a sub 1% level. The self-shielding within the microspheres is found to be responsible for up to 6% reduction of the reconstructed dose in low-gradient regions and must be taken into account. The contribution of trace amounts of other beta+ emitters introduced during the manufacturing process is also discussed. Conclusion: The accelerating pace of clinical adoption of PET-based post-treatment dosimetry is mainly due to advances in both quantitative PET imaging and physical models of dose deposition. We show that the overall physical dose uncertainty in the convolution step can be further reduced, thus making ongoing multi-institutional dose-outcome studies even more reliable.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2015

MO-FG-204-09: High Spatial Resolution and Artifact-Free CT Bone Imaging at Off-Centered Positions: An Application of Model-Based Iterative Reconstruction

Purpose: Although the anatomy of interest should be positioned as close as possible to the isocenter of CT scanners, off-centering may be inevitable during certain exams in clinical practice such as lumbar spine and elbow imaging. Off-centering degrades image sharpness, generates streak artifacts, and sometimes creates blooming artifacts due to truncation. The purpose of this work was to investigate whether the use of model-based image reconstruction (MBIR) can alleviate the negative impacts of off-centering to achieve high quality CT bone imaging. Methods: Both an anthropomorphic phantom and an ex vivo swine elbow sample were scanned at centered and off-centered positions using clinical CT bone scan protocols. The magnitude of off-centering was determined from localizer radiographs. Both FBP and MBIR reconstructions were performed. For FBP, both standard and Bone Plus kernels commonly used in bone imaging were used. Objective assessment of image sharpness, noise standard deviation, and noise nonuniformity were performed. Additionally, we retrospectively analyzed human subject data acquired under off-centered conditions as a validation study. Results: In FBP images of the phantom, off-centering by 10 cm led to a 14% increase in noise (p<1e-3) and a 68% increase in noise nonuniformity (p<0.02). A visible drop in bone sharpness was observed. In contrast, no significant difference in the noise magnitude or the noise nonuniformity between the centered and off-centered MBIR images was found. The image sharpness of off-centered MBIR images outperformed that of FBP images reconstructed with the Bone Plus kernel. In images of the swine elbow off-centered by 20 cm, not only was the noise and spatial resolution performance improved by MBIR, truncation artifacts were also elliminated. The human subject study generated similar results, in which the visibility of the off-centered lumbar spine was significantly improved. Conclusion: High quality CT bone imaging at off-centered positions can be achieved using MBIR. This work was partially supported by an NIH grant R01CA169331 and GE Healthcare. K. Li, D. Gomez-Cardona: Nothing to disclose. G.-H. Chen: Research funded, GE Healthcare; Research funded, Siemens AX. A. Budde, J. Hsieh: Employee, GE Healthcare.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2015

WE-AB-303-09: Rapid Projection Computations for On-Board Digital Tomosynthesis in Radiation Therapy

Purpose: To facilitate fast and accurate iterative volumetric image reconstruction from limited-angle on-board projections. Methods: Intrafraction motion hinders the clinical applicability of modern radiotherapy techniques, such as lung stereotactic body radiation therapy (SBRT). The LIVE system may impact clinical practice by recovering volumetric information via Digital Tomosynthesis (DTS), thus entailing low time and radiation dose for image acquisition during treatment. The DTS is estimated as a deformation of prior CT via iterative registration with on-board images; this shifts the challenge to the computational domain, owing largely to repeated projection computations across iterations. We address this issue by composing efficient digital projection operators from their constituent parts. This allows us to separate the static (projection geometry) and dynamic (volume/image data) parts of projection operations by means of pre-computations, enabling fast on-board processing, while also relaxing constraints on underlying numerical models (e.g. regridding interpolation kernels). Further decoupling the projectors into simpler ones ensures the incurred memory overhead remains low, within the capacity of a single GPU. These operators depend only on the treatment plan and may be reused across iterations and patients. The dynamic processing load is kept to a minimum and maps well to the GPU computational model. Results: We have integrated efficient, pre-computable modules for volumetric ray-casting and FDK-based back-projection with the LIVE processing pipeline. Our results show a 60x acceleration of the DTS computations, compared to the previous version, using a single GPU; presently, reconstruction is attained within a couple of minutes. The present implementation allows for significant flexibility in terms of the numerical and operational projection model; we are investigating the benefit of further optimizations and accurate digital projection sub-kernels. Conclusion: Composable projection operators constitute a versatile research tool which can greatly accelerate iterative registration algorithms and may be conducive to the clinical applicability of LIVE. National Institutes of Health Grant No. R01-CA184173; GPU donation by NVIDIA Corporation.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2015

WE-EF-BRA-07: High Performance Preclinical Irradiation Through Optimized Dual Focal Spot Dose Painting and Online Virtual Isocenter Radiation Field Targeting

Purpose: Advances in radiotherapy practice facilitated by collimation systems to shape radiation fields and image guidance to target these conformal beams have motivated proposals for more complex dose patterns to improve the therapeutic ratio. Recent progress in small animal radiotherapy platforms has provided the foundation to validate the efficacy of such interventions, but robustly delivering heterogeneous dose distributions at the scale and accuracy demanded by preclinical studies remains challenging. This work proposes a dual focal spot optimization method to paint spatially heterogeneous dose regions and an online virtual isocenter targeting method to accurately target the dose distributions. Methods: Two-dimensional dose kernels were empirically measured for the 1 mm diameter circular collimator with radiochromic film in a solid water phantom for the small and large x-ray focal spots on the X-RAD 225Cx microirradiator. These kernels were used in an optimization framework which determined a set of animal stage positions, beam-on times, and focal spot settings to optimally deliver a given desired dose distribution. An online method was developed which defined a virtual treatment isocenter based on a single image projection of the collimated radiation field. The method was demonstrated by optimization of a 6 mm circular 2 Gy target adjoining a 4 mm semicircular avoidance region. Results: The dual focal spot technique improved the optimized dose distribution with the proportion of avoidance region receiving more than 0.5 Gy reduced by 40% compared to the large focal spot technique. Targeting tests performed by irradiating ball bearing targets on radiochromic film pieced revealed the online targeting method improved the three-dimensional accuracy from 0.48 mm to 0.15 mm. Conclusion: The dual focal spot optimization and online virtual isocenter targeting framework is a robust option for delivering dose at the preclinical level and provides a new experimental option for unique radiobiological investigations This work is supported, in part, by the Natural Sciences and Engineering Research Council of Canada and a Mitacs-Accelerate fellowship. P.E. Lindsay, and D.A. Jaffray are listed as inventors of the system described herein. This system has been licensed to Precision X-Ray Inc. for commercial development.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2016

SU-D-207B-07: Development of a CT-Radiomics Based Early Response Prediction Model During Delivery of Chemoradiation Therapy for Pancreatic Cancer

Purpose: Pilot study developing a CT-texture based model for early assessment of treatment response during the delivery of chemoradiation therapy (CRT) for pancreatic cancer. Methods: Daily CT data acquired for 24 pancreatic head cancer patients using CT-on-rails, during the routine CT-guided CRT delivery with a radiation dose of 50.4 Gy in 28 fractions, were analyzed. The pancreas head was contoured on each daily CT. Texture analysis was performed within the pancreas head contour using a research tool (IBEX). Over 1300 texture metrics including: grey level co-occurrence, run-length, histogram, neighborhood intensity difference, and geometrical shape features were calculated for each daily CT. Metric-trend information was established by finding the best fit of either a linear, quadratic, or exponential function for each metric value verses accumulated dose. Thus all the daily CT texture information was consolidated into a best-fit trend type for a given patient and texture metric. Linear correlation was performed between the patient histological response vector (good, medium, poor) and all combinations of 23 patient subgroups (statistical jackknife) determining which metrics were most correlated to response and repeatedly reliable across most patients. Control correlations against CT scanner, reconstruction kernel, and gated/nongated CT images were also calculated. Euclidean distance measure was used to group/sort patient vectors based on the data of these trend-response metrics. Results: We found four specific trend-metrics (Gray Level Coocurence Matrix311-1InverseDiffMomentNorm, Gray Level Coocurence Matrix311-1InverseDiffNorm, Gray Level Coocurence Matrix311-1 Homogeneity2, and Intensity Direct Local StdMean) that were highly correlated with patient response and repeatedly reliable. Our four trend-metric model successfully ordered our pilot response dataset (p=0.00070). We found no significant correlation to our control parameters: gating (p=0.7717), scanner (p=0.9741), and kernel (p=0.8586). Conclusion: We have successfully created a CT-texture based early treatment response prediction model using the CTs acquired during the delivery of chemoradiation therapy for pancreatic cancer. Future testing is required to validate the model with more patient data.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2016

SU-F-T-620: Development of a Convolution/Superposition Dose Engine for CyberKnife System

Purpose: Current CyberKnife treatment planning system (TPS) provided two dose calculation algorithms: Ray-tracing and Monte Carlo. Ray-tracing algorithm is fast, but less accurate, and also can’t handle irregular fields since a multi-leaf collimator system was recently introduced to CyberKnife M6 system. Monte Carlo method has well-known accuracy, but the current version still takes a long time to finish dose calculations. The purpose of this paper is to develop a GPU-based fast C/S dose engine for CyberKnife system to achieve both accuracy and efficiency. Methods: The TERMA distribution from a poly-energetic source was calculated based on beam’s eye view coordinate system, which is GPU friendly and has linear complexity. The dose distribution was then computed by inversely collecting the energy depositions from all TERMA points along 192 collapsed-cone directions. EGSnrc user code was used to pre-calculate energy deposition kernels (EDKs) for a series of mono-energy photons The energy spectrum was reconstructed based on measured tissue maximum ratio (TMR) curve, the TERMA averaged cumulative kernels was then calculated. Beam hardening parameters and intensity profiles were optimized based on measurement data from CyberKnife system. Results: The difference between measured and calculated TMR are less than 1% for all collimators except in the build-up regions. The calculated profiles also showed good agreements with the measured doses within 1% except in the penumbra regions. The developed C/S dose engine was also used to evaluate four clinical CyberKnife treatment plans, the results showed a better dose calculation accuracy than Ray-tracing algorithm compared with Monte Carlo method for heterogeneous cases. For the dose calculation time, it takes about several seconds for one beam depends on collimator size and dose calculation grids. Conclusion: A GPU-based C/S dose engine has been developed for CyberKnife system, which was proven to be efficient and accurate for clinical purpose, and can be easily implemented in TPS.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2016

TU-AB-BRC-06: Dose Calculation in Curved Space

Purpose: Microbeam Radiation Therapy is a preclinical method in radiation oncology that modulates radiation fields on a micrometre scale. Dose calculation is challenging due to arising dose gradients and therapeutically important dose ranges. Monte Carlo (MC) simulations, often used as gold standard, are computationally expensive and hence too slow for the optimisation of treatment parameters in future clinical applications. On the other hand, conventional kernel based dose calculation leads to inaccurate results close to material interfaces. The purpose of this work is to overcome these inaccuracies while keeping computation times low. Methods: A point kernel superposition algorithm is modified to account for tissue inhomogeneities. Instead of conventional ray tracing approaches, methods from differential geometry are applied and the space around the primary photon interaction is locally warped. The performance of this approach is compared to MC simulations and a simple convolution algorithm (CA) for two different phantoms and photon spectra. Results: While peak doses of all dose calculation methods agreed within less than 4% deviations, the proposed approach surpassed a simple convolution algorithm in accuracy by a factor of up to 3 in the scatter dose. In a treatment geometry similar to possible future clinical situations differences between Monte Carlo and the differential geometry algorithm were less than 3%. At the same time the calculation time did not exceed 15 minutes. Conclusion: With the developed method it was possible to improve the dose calculation based on the CA method with respect to accuracy especially at sharp tissue boundaries. While the calculation is more extensive than for the CA method and depends on field size, the typical calculation time for a 20×20 mm{sup 2} field on a 3.4 GHz and 8 GByte RAM processor remained below 15 minutes. Parallelisation and optimisation of the algorithm could lead to further significant calculation time reductions.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2014

Resummation with Wilson lines off the light cone

I review the resummation formalism for organizing large logarithms in perturbative expansion of collinear subprocesses through the variation of Wilson lines off the light cone. A master equation is derived, which involves the evolution kernel resulting from this variation. It is then demonstrated that all the known single- and double-logarithm summations for a parton distribution function or a transverse-momentum-dependent parton distribution can be reproduced from the master equation by applying appropriate soft-gluon approximations to the evolution kernel. Moreover, jet substructures, information which is crucial for particle identification at the Large Hadron Collider and usually acquired from event generators, can also be calculated in this formalism.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2021

Data-Driven Learning for the Mori--Zwanzig Formalism: A Generalization of the Koopman Learning Framework

A theoretical framework which unifies the conventional Mori--Zwanzig formalism and the approximate Koopman learning of deterministic dynamical systems from noiseless observation is presented. In this framework, the Mori--Zwanzig formalism, developed in statistical mechanics to tackle the hard problem of construction of reduced-order dynamics for high-dimensional dynamical systems, can be considered as a natural generalization of the Koopman description of the dynamical system. We next show that, similar to the approximate Koopman learning methods, data-driven methods can be developed for the Mori--Zwanzig formalism with Mori's linear projection operator. We have developed two algorithms to extract the key operators, the Markov and the memory kernel, using time series of a reduced set of observables in a dynamical system. We have adopted the Lorenz `96 system as a test problem and solved for the above operators. These operators exhibit complex behaviors, which are unlikely to be captured by traditional modeling approaches in Mori--Zwanzig analysis. The nontrivial generalized fluctuation-dissipation relationship, which relates the memory kernel with the two-time correlation statistics of the orthogonal dynamics, was numerically verified as a validation of the solved operators. Here we present numerical evidence that the generalized Langevin equation, a key construct in the Mori--Zwanzig formalism, is more advantageous in predicting the evolution of the reduced set of observables than the conventional approximate Koopman operators.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2022

A Framework for Error-Bounded Approximate Computing, with an Application to Dot Products

Approximate computing techniques, which trade off the computation accuracy of an algorithm for better performance and energy efficiency, have been successful in reducing computation and power costs in several domains. However, error sensitive applications in high-performance computing are unable to benefit from existing approximate computing strategies that are not developed with guaranteed error bounds. While approximate computing techniques can be developed for individual high-performance computing applications by domain specialists, this often requires additional theoretical analysis and potentially extensive software modification. Hence, the development of low-level error-bounded approximate computing strategies that can be introduced into any high-performance computing application without requiring additional analysis or significant software alterations is desirable. In this paper, we provide a contribution in this direction by proposing a general framework for designing error-bounded approximate computing strategies and apply it to the dot product kernel to develop \bf qdot---an error-bounded approximate dot product kernel. Following the introduction of qdot, here we perform a theoretical analysis that yields a deterministic bound on the relative approximation error introduced by qdot. Empirical tests are performed to illustrate the tightness of the derived error bound and to demonstrate the effectiveness of qdot on a synthetic dataset, as well as two scientific benchmarks---the conjugate gradient (CG) and power methods. In some instances, using qdot for the dot products in CG can result in many components being quantized to half precision without increasing the iteration count required for convergence to the same solution as CG using a double precision dot product.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2026

Discovery of Probabilistic Dirichlet-to-Neumann Maps on Graphs

Dirichlet-to-Neumann maps enable the coupling of multiphysics simulations across computational subdomains by ensuring continuity of state variables and fluxes at artificial interfaces. We present a novel method for learning Dirichlet-to-Neumann maps on graphs using Gaussian processes, specifically for problems where the data obey a conservation law arising from an underlying partial differential equation. Our approach combines discrete exterior calculus and nonlinear optimal recovery to infer relationships between vertex and edge values. This framework yields data-driven predictions with uncertainty quantification across the entire graph, even when observations are limited to a subset of vertices and edges. By minimizing the reproducing kernel Hilbert space norm while penalizing kernel complexity through maximum likelihood estimation, our method ensures that the resulting surrogate strictly enforces conservation laws without overfitting. We demonstrate our method on two representative applications: subsurface flow in fracture networks and arterial blood flow. Finally, the results demonstrate that the method maintains high accuracy and well-calibrated uncertainty estimates even under severe data scarcity, highlighting its potential for scientific applications where limited data and reliable uncertainty quantification are critical.

Dirichlet-to-Neumann map↗
DOE OSTI2015

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↗
DOE OSTI2016

Delayed Update Algorithms for Quantum Monte Carlo Simulation on GPU: Extended Abstract

QMCPACK is open source scientific software designed to perform Quantum Monte Carlo (QMC) simulation, a first-principles method for describing many-body systems. The evaluation of each Monte Carlo move requires finding the determinant of a dense matrix in the wave functions. This calculation forms a key computational kernel in QMCPACK. After each accepted event, the wave function matrix undergoes a rank-one update to represent a single particle move within the system. The matrix inverse is updated via the Sherman-Morrison formula; occasionally, the explicit inverse must be recomputed to maintain numerical stability.; We propose an alternate approach to this kernel that utilizes QR factorization to maintain stability without refactorization. In addition, algorithms based on a novel delayed update scheme are explored in this effort. This strategy enables probability evaluation for multiple successive Monte Carlo moves, with application of accepted moves to the wave function matrix delayed until an event is denied, or a predetermined limit on acceptances p is reached. Accepted events grouped in this manner are then applied to the matrix en bloc with enhanced arithmetic intensity and computational efficiency. Delayed update Sherman-Morrison probability evaluation algorithms with CPU and GPU-accelerated implementations are tested, profiled, and analyzed. A QR based delayed update algorithm remains in development. Results are evaluated against existing methods for numerical stability and efficiency; emphasis is placed on large systems, for which acceleration is critical.

Computer Science↗
DOE OSTI2018

Evaluating Support for OpenMP Offload Features

The OpenMP language features have been evolving to meet the rapid development in hardware platforms. DOE applications tend to push the bleeding edge of features ratified in the OpenMP specification and tend to expose the rough edges of the features' implementations. The software harness on DOE supercomputers such as Titan and (upcoming) Summit include Cray, Clang, Flang, XL and GCC compilers. It is critical, especially for Summit, that the compilers support OpenMP offloading features. This paper focuses on evaluating support for OpenMP 4.5 target offload directives across compiler implementations on Titan and Summitdev, an early access system, which is one generation removed from Summit's architecture enabling application teams to test the systems' architecture. Our tests not only evaluate the OpenMP implementations but also expose ambiguities in the OpenMP 4.5 specification. We also evaluate compiler implementations using kernels extracted from production DOE applications. This helps in assessing the interaction of different OpenMP directives independent of other application artifacts. We are aware that the implementations are constantly evolving and are advertised as having only partial OpenMP 4.x support. We see this as a synergistic effort to help identify and correct features that are required by DOE applications and prevent deployment delays later on. Going forward, we also plan to interact with standard benchmarking bodies like SPEC/HPG to donate our tests and mini-apps/kernels for potential inclusion in the next release versions of SPEC OMP and SPEC ACCEL benchmark suites.

Monsalve Diaz, Jose↗
DOE OSTI2019

Scalable Causal Graph Learning through a Deep Neural Network

Learning the causal graph in a complex system is crucial for knowledge discovery and decision making, yet it remains a challenging problem because of the unknown nonlinear interaction among system components. Most of the existing methods either rely on predefined kernel or data distribution, or they focus simply on the causality between a single target and the remaining system. This work presents a deep neural network for scalable causal graph learning (SCGL) through low-rank approximation. The SCGL model can explore nonlinearity on both temporal and intervariable relationships without any predefined kernel or distribution assumptions. Through low-rank approximation, the noise influence is reduced, and better accuracy and high scalability are achieved. Experiments using synthetic and real-world datasets show that our SCGL algorithm outperforms existing state-of-the-art methods for causal graph learning.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2021

Extending C++ for Heterogeneous Quantum-Classical Computing

In this report we present qcor - a language extension to C++ and compiler implementation that enables heterogeneous quantum-classical programming, compilation, and execution in a single-source context. Our work provides a first-of-its-kind C++ compiler enabling high-level quantum kernel (function) expression in a quantum-language agnostic manner, as well as a hardware-agnostic, retargetable compiler workflow targeting a number of physical and virtual quantum computing backends. qcor leverages novel Clang plugin interfaces and builds upon the XACC system-level quantum programming framework to provide a state-of-the-art integration mechanism for quantum-classical compilation that leverages the best from the community at-large. qcor translates quantum kernels ultimately to the XACC intermediate representation, and provides user-extensible hooks for quantum compilation routines like circuit optimization, analysis, and placement. This work details the overall architecture and compiler workflow for qcor, and provides a number of illuminating programming examples demonstrating its utility for near-term variational tasks, quantum algorithm expression, and feed-forward error correction schemes.

97 MATHEMATICS AND COMPUTING↗