Search NASA⌕ Search

SEARCH · Search NASA

Results for “Accelerators, Computing”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 631 records · Page 35

TU-AB-303-08: GPU-Based Software Platform for Efficient Image-Guided Adaptive Radiation Therapy

Purpose: In this study, we develop an integrated software platform for adaptive radiation therapy (ART) that combines fast and accurate image registration, segmentation, and dose computation/accumulation methods. Methods: The proposed system consists of three key components; 1) deformable image registration (DIR), 2) automatic segmentation, and 3) dose computation/accumulation. The computationally intensive modules including DIR and dose computation have been implemented on a graphics processing unit (GPU). All required patient-specific data including the planning CT (pCT) with contours, daily cone-beam CTs, and treatment plan are automatically queried and retrieved from their own databases. To improve the accuracy of DIR between pCT and CBCTs, we use the double force demons DIR algorithm in combination with iterative CBCT intensity correction by local intensity histogram matching. Segmentation of daily CBCT is then obtained by propagating contours from the pCT. Daily dose delivered to the patient is computed on the registered pCT by a GPU-accelerated superposition/convolution algorithm. Finally, computed daily doses are accumulated to show the total delivered dose to date. Results: Since the accuracy of DIR critically affects the quality of the other processes, we first evaluated our DIR method on eight head-and-neck cancer cases and compared its performance. Normalized mutual-information (NMI) and normalized cross-correlation (NCC) computed as similarity measures, and our method produced overall NMI of 0.663 and NCC of 0.987, outperforming conventional methods by 3.8% and 1.9%, respectively. Experimental results show that our registration method is more consistent and roust than existing algorithms, and also computationally efficient. Computation time at each fraction took around one minute (30–50 seconds for registration and 15–25 seconds for dose computation). Conclusion: We developed an integrated GPU-accelerated software platform that enables accurate and efficient DIR, auto-segmentation, and dose computation, thus supporting an efficient ART workflow. This work was supported by NIH/NCI under grant R42CA137886.

60 APPLIED LIFE SCIENCES↗

GPU-accelerated denoising of 3D magnetic resonance images

The raw computational power of GPU accelerators enables fast denoising of 3D MR images using bilateral filtering, anisotropic diffusion, and non-local means. In practice, applying these filtering operations requires setting multiple parameters. This study was designed to provide better guidance to practitioners for choosing the most appropriate parameters by answering two questions: what parameters yield the best denoising results in practice? And what tuning is necessary to achieve optimal performance on a modern GPU? To answer the first question, we use two different metrics, mean squared error (MSE) and mean structural similarity (MSSIM), to compare denoising quality against a reference image. Surprisingly, the best improvement in structural similarity with the bilateral filter is achieved with a small stencil size that lies within the range of real-time execution on an NVIDIA Tesla M2050 GPU. Moreover, inappropriate choices for parameters, especially scaling parameters, can yield very poor denoising performance. To answer the second question, we perform an autotuning study to empirically determine optimal memory tiling on the GPU. The variation in these results suggests that such tuning is an essential step in achieving real-time performance. These results have important implications for the real-time application of denoising to MR images in clinical settings that require fast turn-around times.

97 MATHEMATICS AND COMPUTING↗

Computational Study of 15 MeV Linac Radiation Doses at the UNLV Accelerator Facility

Linear electron accelerators (linacs) are used to generate Bremsstrahlung photons through the interaction of an electron beam with a heavy metal target. These Bremsstrahlung photons with endpoint energy equal to that of the impinging electron beam are typically collimated into a desired beam shape (fan or cone beams). At electron energies larger than 10 MeV, the interaction of the newly produced photons with the linac component materials leads to the production of neutrons. These neutrons are often undesired and pollute the collimated photon beam while increasing the dose rates at the facility due to the linac operation. MCNPX 2.7.0 code was used to investigate the production of neutrons due to the normal operation of a 15 MeV linac. Dose rates within the UNLV accelerator facility were calculated for both photons and neutrons. Furthermore, the irradiation of several materials using the K15 linac was modeled; the resulting neutron fluxes and dose rates were determined. (authors)

43 PARTICLE ACCELERATORS↗

Complete quasilinear model for the acceleration-driven lower hybrid drift instability and a computational assessment of its validity

A complete quasilinear model is derived for the electrostatic acceleration-driven lower hybrid drift instability in a uniform two-species low-beta plasma in which current is perpendicular to the background magnetic field. The model consists of coupled nonlinear velocity space diffusion equations for the volume-averaged ion and electron distribution functions. Each species' diffusion coefficient depends on a time-evolving spectral density of the electric-field energy per unit volume and a time-evolving dispersion relation. The dispersion relation is expressed analytically in integral form without the use of asymptotic limits and applies to arbitrary distribution functions, so long as they can be expressed as a function of one velocity coordinate, e.g., f⁡(vy) or f⁡(v⊥). The quasilinear model conserves energy and is complete in that it fully describes the evolution of the distribution functions, including resonant and nonresonant particle-wave interactions, while accounting for distribution-function-dependent mixed-complex frequencies. Further, the quasilinear diffusion model is solved numerically and self-consistently using a Crank-Nicolson temporal discretization and a second-order finite-volume velocity-space discretization. Numerical solutions are compared to nonlinear fourth-order accurate continuum kinetic Vlasov-Poisson simulations. Evolution of electric-field energy, growth rates, distribution functions, and diffusion coefficients are shown to be in agreement with Vlasov simulations. The quasilinear model is shown to predict anomalous transport terms, like resistivity and heating, to within a factor of order unity. Discrepancies between the quasilinear model and Vlasov simulations are assessed and attributed primarily to lack of damping in the quasilinear description and to the use of unperturbed-orbit susceptibilities in the linear theory dispersion relation. The results illuminate the predictive accuracy of the quasilinear model, place approximate bounds on its validity, and provide much needed vetting of quasilinear theory's ability to predict the nonlinear state of a microturbulent plasma.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Computing chemical potentials with machine-learning-accelerated simulations to accurately predict thermodynamic properties of molten salts

The successful design and deployment of next-generation nuclear technologies heavily rely on thermodynamic data for relevant molten salt systems. However, the lack of accurate force fields and efficient methods has limited the quality of thermodynamic predictions from atomistic simulations. Here we propose an efficient free energy framework for computing chemical potentials, which is the central free energy quantity behind many thermodynamic properties. We accelerate our simulations without sacrificing accuracy by using machine learning interatomic potentials trained on density functional theory (DFT) data. Using lithium chloride as our model system, we compute chemical potentials with DFT-accuracy for solid and liquid phases by transmuting ions into noninteracting particles. Notably, in the liquid phase, we demonstrate consistency whether we transmute one ion pair or the entire system into ideal gas particles. By locating the temperature where the chemical potential of solid and liquid phases cross, we predict a melting point of 880 ± 18 K for lithium chloride, which is remarkably close to the experimental value of 883 K. With this successful demonstration, we lay the foundation for high-throughput thermodynamic predictions of many properties that can be derived from the chemical potentials of the minority and majority components in molten salts.

Gibson, Luke D. [Oak Ridge National Laboratory (OR↗

DFT-FE 1.0: A massively parallel hybrid CPU-GPU density functional theory code using finite-element discretization

In this work, we present DFT-FE 1.0, building on DFT-FE 0.6 [Comput. Phys. Commun. 246, 106853 (2020)], to conduct fast and accurate large-scale density functional theory (DFT) calculations (reaching ~ 100,000 electrons) on both many-core CPU and hybrid CPU-GPU computing architectures. This work involves improvements in the real-space formulation—via an improved treatment of the electrostatic interactions that substantially enhances the computational efficiency—as well high-performance computing aspects, including the GPU acceleration of all the key compute kernels in DFT-FE. We demonstrate the accuracy by comparing the ground-state energies, ionic forces and cell stresses on a wide-range of benchmark systems against those obtained from widely used DFT codes. Further, we demonstrate the numerical efficiency of our implementation, which yields ~ 20× CPU-GPU speed-up by using GPU acceleration on hybrid CPU-GPU nodes. Notably, owing to the parallel-scaling of the GPU implementation, we obtain wall-times of 80–140 seconds for full ground-state calculations, with stringent accuracy, on benchmark systems containing ~ 6, 000 – 15,000 electrons.

pseudopotential↗

Accelerating materials science with high-throughput computations and machine learning

With unprecedented amounts of materials data generated from experiments as well as high-throughput density functional theory calculations, machine learning techniques has the potential to greatly accelerate materials discovery and design. In this report we review our efforts in the Materials Virtual Lab to integrate software automation, data generation and curation and machine learning to (i) design and optimize technological materials for energy storage, energy efficiency and high-temperature alloys; (ii) develop scalable quantum-accurate models, and (iii) enhance the speed and accuracy in interpreting characterization spectra.

36 MATERIALS SCIENCE↗

Four-dimensional volume-of-interest reconstruction for cone-beam computed tomography-guided radiation therapy

Purpose: Data sufficiency are a major problem in four-dimensional cone-beam computed tomography (4D-CBCT) on linear accelerator-integrated scanners for image-guided radiotherapy. Scan times must be in the range of 4-6 min to avoid undersampling artifacts. Various image reconstruction algorithms have been proposed to accommodate undersampled data acquisitions, but these algorithms are computationally expensive, may require long reconstruction times, and may require algorithm parameters to be optimized. The authors present a novel reconstruction method, 4D volume-of-interest (4D-VOI) reconstruction which suppresses undersampling artifacts and resolves lung tumor motion for undersampled 1-min scans. The 4D-VOI reconstruction is much less computationally expensive than other 4D-CBCT algorithms. Methods: The 4D-VOI method uses respiration-correlated projection data to reconstruct a four-dimensional (4D) image inside a VOI containing the moving tumor, and uncorrelated projection data to reconstruct a three-dimensional (3D) image outside the VOI. Anatomical motion is resolved inside the VOI and blurred outside the VOI. The authors acquired a 1-min. scan of an anthropomorphic chest phantom containing a moving water-filled sphere. The authors also used previously acquired 1-min scans for two lung cancer patients who had received CBCT-guided radiation therapy. The same raw data were used to test and compare the 4D-VOI reconstruction with the standard 4D reconstruction and the McKinnon-Bates (MB) reconstruction algorithms. Results: Both the 4D-VOI and the MB reconstructions suppress nearly all the streak artifacts compared with the standard 4D reconstruction, but the 4D-VOI has 3-8 times greater contrast-to-noise ratio than the MB reconstruction. In the dynamic chest phantom study, the 4D-VOI and the standard 4D reconstructions both resolved a moving sphere with an 18 mm displacement. The 4D-VOI reconstruction shows a motion blur of only 3 mm, whereas the MB reconstruction shows a motion blur of 13 mm. With graphics processing unit hardware used to accelerate computations, the 4D-VOI reconstruction required a 40-s reconstruction time. Conclusions: 4D-VOI reconstruction effectively reduces undersampling artifacts and resolves lung tumor motion in 4D-CBCT. The 4D-VOI reconstruction is computationally inexpensive compared with more sophisticated iterative algorithms. Compared with these algorithms, our 4D-VOI reconstruction is an attractive alternative in 4D-CBCT for reconstructing target motion without generating numerous streak artifacts.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

Accelerating Development of EV Batteries Through Computer-Aided Engineering (Presentation)

The Department of Energy's Vehicle Technology Program has launched the Computer-Aided Engineering for Automotive Batteries (CAEBAT) project to work with national labs, industry and software venders to develop sophisticated software. As coordinator, NREL has teamed with a number of companies to help improve and accelerate battery design and production. This presentation provides an overview of CAEBAT, including its predictive computer simulation of Li-ion batteries known as the Multi-Scale Multi-Dimensional (MSMD) model framework. MSMD's modular, flexible architecture connects the physics of battery charge/discharge processes, thermal control, safety and reliability in a computationally efficient manner. This allows independent development of submodels at the cell and pack levels.

25 ENERGY STORAGE↗

TU-FG-201-04: Computer Vision in Autonomous Quality Assurance of Linear Accelerators

Purpose: Routine quality assurance (QA) of linear accelerators represents a critical and costly element of a radiation oncology center. Recently, a system was developed to autonomously perform routine quality assurance on linear accelerators. The purpose of this work is to extend this system and contribute computer vision techniques for obtaining quantitative measurements for a monthly multi-leaf collimator (MLC) QA test specified by TG-142, namely leaf position accuracy, and demonstrate extensibility for additional routines. Methods: Grayscale images of a picket fence delivery on a radioluminescent phosphor coated phantom are captured using a CMOS camera. Collected images are processed to correct for camera distortions, rotation and alignment, reduce noise, and enhance contrast. The location of each MLC leaf is determined through logistic fitting and a priori modeling based on knowledge of the delivered beams. Using the data collected and the criteria from TG-142, a decision is made on whether or not the leaf position accuracy of the MLC passes or fails. Results: The locations of all MLC leaf edges are found for three different picket fence images in a picket fence routine to 0.1mm/1pixel precision. The program to correct for image alignment and determination of leaf positions requires a runtime of 21– 25 seconds for a single picket, and 44 – 46 seconds for a group of three pickets on a standard workstation CPU, 2.2 GHz Intel Core i7. Conclusion: MLC leaf edges were successfully found using techniques in computer vision. With the addition of computer vision techniques to the previously described autonomous QA system, the system is able to quickly perform complete QA routines with minimal human contribution.

60 APPLIED LIFE SCIENCES↗

ACE3P Computations of Wakefield Coupling in the CLIC Two-Beam Accelerator

The Compact Linear Collider (CLIC) provides a path to a multi-TeV accelerator to explore the energy frontier of High Energy Physics. Its novel two-beam accelerator concept envisions rf power transfer to the accelerating structures from a separate high-current decelerator beam line consisting of power extraction and transfer structures (PETS). It is critical to numerically verify the fundamental and higher-order mode properties in and between the two beam lines with high accuracy and confidence. To solve these large-scale problems, SLAC's parallel finite element electromagnetic code suite ACE3P is employed. Using curvilinear conformal meshes and higher-order finite element vector basis functions, unprecedented accuracy and computational efficiency are achieved, enabling high-fidelity modeling of complex detuned structures such as the CLIC TD24 accelerating structure. In this paper, time-domain simulations of wakefield coupling effects in the combined system of PETS and the TD24 structures are presented. The results will help to identify potential issues and provide new insights on the design, leading to further improvements on the novel CLIC two-beam accelerator scheme.

43 PARTICLE ACCELERATORS↗

DIFFUSIVE PARTICLE ACCELERATION IN SHOCKED, VISCOUS ACCRETION DISKS: GREEN'S FUNCTION ENERGY DISTRIBUTION

The acceleration of relativistic particles in a viscous accretion disk containing a standing shock is investigated as a possible explanation for the energetic outflows observed around radio-loud black holes. The energy/space distribution of the accelerated particles is computed by solving a transport equation that includes the effects of first-order Fermi acceleration, bulk advection, spatial diffusion, and particle escape. The velocity profile of the accreting gas is described using a model for shocked viscous disks recently developed by the authors, and the corresponding Green's function distribution for the accelerated particles in the disk and the outflow is obtained using a classical method based on eigenfunction analysis. The accretion-driven, diffusive shock acceleration scenario explored here is conceptually similar to the standard model for the acceleration of cosmic rays at supernova-driven shocks. However, in the disk application, the distribution of the accelerated particles is much harder than would be expected for a plane-parallel shock with the same compression ratio. Hence the disk environment plays a key role in enhancing the efficiency of the shock acceleration process. The presence of the shock helps to stabilize the disk by reducing the Bernoulli parameter, while channeling the excess binding energy into the escaping relativistic particles. In applications to M87 and Sgr A*, we find that the kinetic power in the jet is {approx}0.01 M-dot c{sup 2}, and the outflowing relativistic particles have a mean energy {approx}300 times larger than that of the thermal gas in the disk at the shock radius. Our results suggest that a standing shock may be an essential ingredient in accretion onto underfed black holes, helping to resolve the long-standing problem of the stability of advection-dominated accretion disks.

79 ASTRONOMY AND ASTROPHYSICS↗

Geometry control in adaptive truss structures

Forward and inverse kinematics equations are derived for the large geometry maneuver of adaptive trusses. A new algorithm based on higher-order multistep methods is proposed as a means of computing the length control for a described large geometry maneuver. The algorithm is shown to improve in computational speed at least five times over the algorithm presented in an earlier paper. The acceleration in control computation and other features such as varying velocity profiles and curved trajectories are illustrated by simulation results.

Ramesh, A. V.↗

Flexible silicon photonic architecture for accelerating distributed deep learning

The increasing size and complexity of deep learning (DL) models have led to the wide adoption of distributed training methods in datacenters (DCs) and high-performance computing (HPC) systems. However, communication among distributed computing units (CUs) has emerged as a major bottleneck in the training process. In this study, we propose Flex-SiPAC, a flexible silicon photonic accelerated compute cluster designed to accelerate multi-tenant distributed DL training workloads. Flex-SiPAC takes a co-design approach that combines a silicon photonic hardware platform with a tailored collective algorithm, optimized to leverage the unique physical properties of the architecture. The hardware platform integrates a novel wavelength-reconfigurable transceiver design and a micro-resonator-based wavelength-reconfigurable switch, enabling the system to achieve flexible bandwidth steering in the wavelength domain. The collective algorithm is designed to support reconfigurable topologies, enabling efficient all-reduce communications that are commonly used in DL training. The feasibility of the Flex-SiPAC architecture is demonstrated through two testbed experiments. First, an optical testbed experiment demonstrates the flexible routing of wavelengths by shuffling an array of input wavelengths using a custom-designed spatial-wavelength selective switch. Second, a four-GPU testbed running two DL workloads shows a 23% improvement in job completion time compared to a similarly sized leaf-spine topology. We further evaluate Flex-SiPAC using large-scale simulations, which show that Flex-SiPAC is able to reduce the communication time by 26% to 29% compared to state-of-the-art compute clusters under representative collective operations.

Wu, Zhenguo (ORCID:0000000322847985)↗

Evaluating Portable Parallelization Strategies for Heterogeneous Architectures in High Energy Physics

High-energy physics (HEP) experiments have developed millions of lines of code over decades that are optimized to run on traditional x86 CPU systems. However, we are seeing a rapidly increasing fraction of floating point computing power in leadership-class computing facilities and traditional data centers coming from new accelerator architectures, such as GPUs. HEP experiments are now faced with the untenable prospect of rewriting millions of lines of x86 CPU code, for the increasingly dominant architectures found in these computational accelerators. This task is made more challenging by the architecture-specific languages and APIs promoted by manufacturers such as NVIDIA, Intel and AMD. Producing multiple, architecture-specific implementations is not a viable scenario, given the available person power and code maintenance issues. The Portable Parallelization Strategies team of the HEP Center for Computational Excellence is investigating the use of Kokkos, SYCL, OpenMP, std::execution::parallel and alpaka as potential portability solutions that promise to execute on multiple architectures from the same source code, using representative use cases from major HEP experiments, including the DUNE experiment of the Long Baseline Neutrino Facility, and the ATLAS and CMS experiments of the Large Hadron Collider. This cross-cutting evaluation of portability solutions using real applications will help inform and guide the HEP community when choosing their software and hardware suites for the next generation of experimental frameworks. We present the outcomes of our studies, including performance metrics, porting challenges, API evaluations, and build system integration.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Unsteady Aerodynamic Force Sensing from Measured Strain

A simple approach for computing unsteady aerodynamic forces from simulated measured strain data is proposed in this study. First, the deflection and slope of the structure are computed from the unsteady strain using the two-step approach. Velocities and accelerations of the structure are computed using the autoregressive moving average model, on-line parameter estimator, low-pass filter, and a least-squares curve fitting method together with analytical derivatives with respect to time. Finally, aerodynamic forces over the wing are computed using modal aerodynamic influence coefficient matrices, a rational function approximation, and a time-marching algorithm. A cantilevered rectangular wing built and tested at the NASA Langley Research Center (Hampton, Virginia, USA) in 1959 is used to validate the simple approach. Unsteady aerodynamic forces as well as wing deflections, velocities, accelerations, and strains are computed using the CFL3D computational fluid dynamics (CFD) code and an MSC/NASTRAN code (MSC Software Corporation, Newport Beach, California, USA), and these CFL3D-based results are assumed as measured quantities. Based on the measured strains, wing deflections, velocities, accelerations, and aerodynamic forces are computed using the proposed approach. These computed deflections, velocities, accelerations, and unsteady aerodynamic forces are compared with the CFL3D/NASTRAN-based results. In general, computed aerodynamic forces based on the lifting surface theory in subsonic speeds are in good agreement with the target aerodynamic forces generated using CFL3D code with the Euler equation. Excellent aeroelastic responses are obtained even with unsteady strain data under the signal to noise ratio of -9.8dB. The deflections, velocities, and accelerations at each sensor location are independent of structural and aerodynamic models. Therefore, the distributed strain data together with the current proposed approaches can be used as distributed deflection, velocity, and acceleration sensors. This research demonstrates the feasibility of obtaining induced drag and lift forces through the use of distributed sensor technology with measured strain data. An active induced drag control system thus can be designed using the two computed aerodynamic forces, induced drag and lift, to improve the fuel efficiency of an aircraft. Interpolation elements between structural finite element grids and the CFD grids and centroids are successfully incorporated with the unsteady aeroelastic computation scheme. The most critical technology for the success of the proposed approach is the robust on-line parameter estimator, since the least-squares curve fitting method depends heavily on aeroelastic system frequencies and damping factors.

shape sensing↗

H-mode accelerating structures with PMQ focusing for low-beta ion beams

We are developing high-efficiency normal-conducting RF accelerating structures based on inter-digital H-mode (IH) cavities and the transverse beam focusing with permanent-magnet quadrupoles (PMQ), for beam velocities in the range of a few percent of the speed of light. Such IH-PMQ accelerating structures following a short RFQ can be used in the front end of ion linacs or in stand-alone applications, e.g. a compact deuteron-beam accelerator up to the energy of several MeV. Results of combined 3-D modeling for a full IH-PMQ accelerator tank - electromagnetic computations, beam-dynamics simulations with high currents, and thermal-stress analysis - are presented. The accelerating field profile in the tank is tuned to provide the best beam propagation using coupled iterations of electromagnetic and beam-dynamics modeling. A cold model of the IH-PMQ tank is being manufactured.

43 PARTICLE ACCELERATORS↗

Predicting Atomistic Transitions with Transformers

Accurate knowledge of the atomistic transition pathways in materials and material surfaces is crucial for many material science problems. However, conventional simulation techniques used to find these transitions are extremely computationally intensive. Even with large-scale, accelerated material simulations, the computational cost constrains the applicable domain in practice. Machine learning models, with the potential to learn the complex emergent behaviors governing atomistic transitions as a fast surrogate model, have great promise to predict transitions with a vastly reduced computational cost. Here, we demonstrate how transformers can be trained to predict atomistic transitions in nano-clusters. We show how we evaluate physical validity of the predictions and how a multitude of additional, different microstates can be generated by slightly varying the data provided to the model.

36 MATERIALS SCIENCE↗