Search NASA⌕ Search

SEARCH · Search NASA

Results for “accelerated 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 451 records · Page 25

A Logical Model of Homology for Comparative Biology

There is a growing body of research on the evolution of anatomy in a wide variety of organisms. Discoveries in this field could be greatly accelerated by computational methods and resources that enable these findings to be compared across different studies and different organisms and linked with the genes responsible for anatomical modifications. Homology is a key concept in comparative anatomy; two important types are historical homology (the similarity of organisms due to common ancestry) and serial homology (the similarity of repeated structures within an organism). We explored how to most effectively represent historical and serial homology across anatomical structures to facilitate computational reasoning. We assembled a collection of homology assertions from the literature with a set of taxon phenotypes for the skeletal elements of vertebrate fins and limbs from the Phenoscape Knowledgebase. Using seven competency questions, we evaluated the reasoning ramifications of two logical models: the Reciprocal Existential Axioms (REA) homology model and the Ancestral Value Axioms (AVA) homology model. The AVA model returned all user-expected results in addition to the search term and any of its subclasses. The AVA model also returns any superclass of the query term in which a homology relationship has been asserted. The REA model returned the user-expected results for five out of seven queries. We identify some challenges of implementing complete homology queries due to limitations of OWL reasoning. This work lays the foundation for homology reasoning to be incorporated into other ontology-based tools, such as those that enable synthetic supermatrix construction and candidate gene discovery.

59 BASIC BIOLOGICAL SCIENCES↗

Polarizable atomic multipole X-ray refinement: application to peptide crystals

A method to accelerate the computation of structure factors from an electron density described by anisotropic and aspherical atomic form factors via fast Fourier transformation is described for the first time. Recent advances in computational chemistry have produced force fields based on a polarizable atomic multipole description of biomolecular electrostatics. In this work, the Atomic Multipole Optimized Energetics for Biomolecular Applications (AMOEBA) force field is applied to restrained refinement of molecular models against X-ray diffraction data from peptide crystals. A new formalism is also developed to compute anisotropic and aspherical structure factors using fast Fourier transformation (FFT) of Cartesian Gaussian multipoles. Relative to direct summation, the FFT approach can give a speedup of more than an order of magnitude for aspherical refinement of ultrahigh-resolution data sets. Use of a sublattice formalism makes the method highly parallelizable. Application of the Cartesian Gaussian multipole scattering model to a series of four peptide crystals using multipole coefficients from the AMOEBA force field demonstrates that AMOEBA systematically underestimates electron density at bond centers. For the trigonal and tetrahedral bonding geometries common in organic chemistry, an atomic multipole expansion through hexadecapole order is required to explain bond electron density. Alternatively, the addition of interatomic scattering (IAS) sites to the AMOEBA-based density captured bonding effects with fewer parameters. For a series of four peptide crystals, the AMOEBA–IAS model lowered R{sub free} by 20–40% relative to the original spherically symmetric scattering model.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗

Security vulnerabilities of connected vehicle streams and their impact on cooperative driving

Autonomous vehicles capable of navigating unpredictable real-world environments with little human feedback are a reality today. Such systems rely heavily on onboard sensors such as cameras, radar/LIDAR, and GPS as well as capabilities such as 3G/4G connectivity and V2V/V2I communication to make real-time maneuvering decisions. Autonomous vehicle control imposes very strict requirements on the security of the communication channels used by the vehicle to exchange information as well as the control logic that performs complex driving tasks such as adapting vehicle velocity or changing lanes. This research presents a first look at the effects of security attacks on the communication channel as well as sensor tampering of a connected vehicle stream equipped to achieve CACC. Our simulation results show that an insider attack can cause significant instability in the CACC vehicle stream. We also outline how different countermeasures, such as downgrading to ACC mode, could potentially be used to improve the security and safety of the connected vehicle streams.

42 ENGINEERING↗

Establish the basis for Breadth-First Search on Frontier System: XBFS on AMD GPUs

Graphics Processing Units (GPUs) offer significant potential for accelerating various computational tasks, including Breadth-First Search (BFS). Numerous efforts have been made to deploy BFS on GPUs effectively. To address the dynamic nature of BFS, XBFS, the state-of-the-art work, employs an adaptive strategy that leverages different optimized frontier queue generation designs, accommodating the varying characteristics of levels in BFS. While XBFS demonstrates excellent performance on NVIDIA Quadro P6000 GPUs, it faces challenges when deployed on AMD GPUs. In this work, we present our efforts to implement XBFS’s adaptive approach on Frontier, the most powerful supercomputer system, by porting XBFS to AMD MI250X GPUs. Through targeted optimizations tailored to the unique features of AMD GPUs, our implementation achieves an average performance of 43 Giga-Traversed Edges Per Second (GTEPS) per Graphics Compute Dies (GCD). Based on these results, we observe potential for surpassing the performance of the official Frontier results from the Graph500 benchmark released in June 2024.

Yang, Haoshen↗

PANTHER: A Programmable Architecture for Neural Network Training Harnessing Energy-Efficient ReRAM

The wide adoption of deep neural networks has been accompanied by ever-increasing energy and performance demands due to the expensive nature of training them. Additionally, numerous special-purpose architectures have been proposed to accelerate training: both digital and hybrid digital-analog using resistive RAM (ReRAM) crossbars. ReRAM-based accelerators have demonstrated the effectiveness of ReRAM crossbars at performing matrix-vector multiplication operations that are prevalent in training. However, they still suffer from inefficiency due to the use of serial reads and writes for performing the weight gradient and update step. A few works have demonstrated the possibility of performing outer products in crossbars, which can be used to realize the weight gradient and update step without the use of serial reads and writes. However, these works have been limited to low precision operations which are not sufficient for typical training workloads. Moreover, they have been confined to a limited set of training algorithms for fully-connected layers only. To address these limitations, we propose a bit-slicing technique for enhancing the precision of ReRAM-based outer products, which is substantially different from bit-slicing for matrix-vector multiplication only. We incorporate this technique into a crossbar architecture with three variants catered to different training algorithms. To evaluate our design on different types of layers in neural networks (fully-connected, convolutional, etc.) and training algorithms, we develop PANTHER, an ISA-programmable training accelerator with compiler support. Our design can also be integrated into other accelerators in the literature to enhance their efficiency. Our evaluation shows that PANTHER achieves up to 8.02×, 54.21×, and 103× energy reductions as well as 7.16×, 4.02×, and 16× execution time reductions compared to digital accelerators, ReRAM-based accelerators, and GPUs, respectively.

42 ENGINEERING↗

SU-E-T-608: Performance Comparison of Four Commercial Treatment Planning Systems Applied to Intensity-Modulated Radiotherapy

Purpose: To compare the performances of four commercial treatment planning systems (TPS) used for the intensity-modulated radiotherapy (IMRT). Methods: Ten patients of nasopharyngeal (4 cases), esophageal (3 cases) and cervical (3 cases) cancer were randomly selected from a 3-month IMRT plan pool at one radiotherapy center. For each patient, four IMRT plans were newly generated by using four commercial TPS (Corvus, Monaco, Pinnacle and Xio), and then verified with Matrixx (two-dimensional array/IBA Company) on Varian23EX accelerator. A pass rate (PR) calculated from the Gamma index by OminiPro IMRT 1.5 software was evaluated at four plan verification standards (1%/1mm, 2%/2mm, 3%/3mm, 4%/4mm and 5%/5mm) for each treatment plan. Overall and multiple pairwise comparisons of PRs were statistically conducted by analysis of covariance (ANOVA) F and LSD tests among four TPSs. Results: Overall significant (p>0.05) differences of PRs were found among four TPSs with F test values of 3.8 (p=0.02), 21.1(>0.01), 14.0 (>0.01), 8.3(>0.01) at standards of 1%/1mm to 4%/4mm respectively, except at 5%/5mm standard with 2.6 (p=0.06). All means (standard deviation) of PRs at 3%/3mm of 94.3 ± 3.3 (Corvus), 98.8 ± 0.8 (Monaco), 97.5± 1.7 (Pinnacle), 98.4 ± 1.0 (Xio) were above 90% and met clinical requirement. Multiple pairwise comparisons had not demonstrated a consistent low or high pattern on either TPS. Conclusion: Matrixx dose verification results show that the validation pass rates of Monaco and Xio plans are relatively higher than those of the other two; Pinnacle plan shows slight higher pass rate than Corvus plan; lowest pass rate was achieved by the Corvus plan among these four kinds of TPS.

60 APPLIED LIFE SCIENCES↗

SU-F-T-80: A Mobile Application for Intra-Operative Electron Radiotherapy Treatment Planning

Purpose: Intraoperative electron radiotherapy (IORT) poses a unique set of challenges for treatment planning. Planning must be performed in a busy operating room environment over a short timeframe often with little advance knowledge of the treatment depth or applicator size. Furthermore, IORT accelerators can have a large number of possible applicators, requiring extensive databooks that must be searched for the appropriate dosimetric parameters. The goal of this work is to develop a software tool to assist in the planning process that is suited to the challenges faced in the IORT environment. Methods: We developed a mobile application using HTML5 and Javascript that can be deployed to tablet devices suitable for use in the operating room. The user selects the desired treatment parameters cone diameter, bevel angle, and energy (a total of 141 datasets) and desired bolus. The application generates an interactive display that allows the user to dynamically select points on the depth-dose curve and to visualize the shape of the corresponding isodose contours. The user can indicate a prescription isodose line or depth. The software performs a monitor unit calculation and generates a PDF report. Results: We present our application, which is now used routinely in our IORT practice. It has been employed successfully in over 23 cases. The interactivity of the isodose distributions was found to be of particular use to physicians who are less-frequent IORT users, as well as for the education of residents and trainees. Conclusion: This software has served as a useful tool in IORT planning, and demonstrates the need for treatment planning tools that are designed for the specialized challenges encountered in IORT. This software is the subject of a license agreement with the IntraOp Medical Corporation. This software is the subject of a license agreement between Massachusetts General Hospital / Partners Healthcare and the IntraOp Medical Corporation. CLW is consulting on software development with the IntraOp Medical Corporation.

60 APPLIED LIFE SCIENCES↗

Highly scalable distributed-memory sparse triangular solution algorithms.

This paper presents a highly efficient distributed-memory parallel sparse triangular solver. The triangular solution phase is often performed following factorization phase in the sparse linear solvers and has become increasingly computa­tionally expensive for direct solvers with many right hand sides (RHSs) or preconditioned iterative solvers. However, the low arithmetic intensity and sequential nature of the triangular solve algorithm pose performance challenges for its large-scale distributed-memory parallelization. In this work, we propose several strategies to enhance scalability of an algorithm with 2D block cyclic process layout. First, an asynchronous binary-tree-based communication scheme implemented via non-blocking MPI functions is leveraged to broadcast partial solutions and reduce partial updates among a subset of processes for each block column and row of the triangular matrix, respectively. This scheme reduces message latency, improves communication load balance and significantly accelerates asynchronous execution of the triangular solve. In addition, efficient BLAS operations and threading implementations are exploited to accelerate local computations and further reduce process idle time. The proposed strategies are implemented in SuperLU_DIST and numerical experiments show up to 4.4x improvement with one right-hand side and up to 6.1x improvement with 50 right- hand sides on 4096 processes, compared to the current release. This is the first time that sparse triangular solution is demonstrated strong scaling on more than 4000 cores.

Liu, Yang↗

MTIP Single Particle Imaging (Spinifel) v0.1.0

Performs Multi-Tiered Iterative Phasing for Single Particle Imaging. The Spinifel code uses distributed computing (using MPI and Legion) to spread work across many compute nodes. On-node computation is accelerated using GPGPUs (using CuPy, and CUDA).

Blaschke, Johannes↗

Ipopt Interface to Re::Solve Linear Solver

The software provides Ipopt optimization package an interface to the Re::Solve linear solver library. Re::Solve features GPU-resident direct and iterative linear solvers that could be used to accelerate optimization computations.

Alam, Maksudul [Oak Ridge National Laboratory (ORN↗

A Global Repository for Planet-Sized Experiments and Observations

Greenland’s ice loss doubled between 1996 and 2005. Nine of the ten warmest years on record have occurred since 2000. Carbon dioxide levels in the air are at their highest in 650,000 years. Are these phenomena related, and if so, what can they tell us about the Earth’s climate in the coming years and decades? The answer will require sophisticated computer models—and a way of dealing with the data they produce. Past climate patterns can be discerned by examining variables such as temperature and precipitation records, lake sediments, tree rings, and ice cores. Projecting future regional and global climate patterns, however, requires computer models that can simulate the complex interplay between physical, chemical, and biological processes occurring in the atmosphere, ocean, and on land and sea ice. These models are extremely computationally demanding. Over the past several decades, an ongoing revolution in high-capacity, high-performance computing has accelerated climate modeling. In 1998, a 1-year simulation of global climate using a 300- kilometer grid could be run in a day. In 2002, it took 4 or 5 hours. In 2015, this same model can be run on a high-end desktop in minutes.

54 ENVIRONMENTAL SCIENCES↗

An overview of the Hadoop/MapReduce/HBase framework and its current applications in bioinformatics

Bioinformatics researchers are increasingly confronted with analysis of ultra large-scale data sets, a problem that will only increase at an alarming rate in coming years. Recent developments in open source software, that is, the Hadoop project and associated software, provide a foundation for scaling to petabyte scale data warehouses on Linux clusters, providing fault-tolerant parallelized analysis on such data using a programming style named MapReduce. An overview is given of the current usage within the bioinformatics community of Hadoop, a top-level Apache Software Foundation project, and of associated open source software projects. The concepts behind Hadoop and the associated HBase project are defined, and current bioinformatics software that employ Hadoop is described. The focus is on next-generation sequencing, as the leading application area to date.

43 PARTICLE ACCELERATORS↗

SPi User Manual V0.1

This document describes the Serial Powering Interface (SPi) ASIC. SPi is a general purpose ASIC prototype designed for use in serial powering of silicon detector instrumentation. This description is written as a user manual to aid application, not as a design description. SPi is a generic custom ASIC, manufactured in 0.25 {mu}m CMOS by TSMC, to interface between a constant current source and silicon detector read-out chips. There is no SEU (single event upset) protection, but most (not all) components are radiation tolerant design. An operating voltage of 1.2 to 2.5 volts and other design features make the IC suitable for a variety of serial powering architectures and ROICs. It should be noted that the device is likely to be a prototype for demonstration rather than a product for inclusion in a detector. The next design(s), SPin, are likely to be designed for a specific application (eg SLHC). The component includes: (1) Seven bi-directional LVDS-like buffers for high data rate links to/from the read-out chips. These are AC coupled (series capacitor) off-chip for DC level conversion; (2) A programmable internal programmable shunt regulator to provide a defined voltage to readout chips when linked in a serial powering chain; (3) A programmable internal shunt regulator control circuit for external transistor control; (4) Shunt current measurement (for internal shunt regulator); (5) A programmable internal shunt regulator current alarm; and (6) Two programmable linear regulators.

43 PARTICLE ACCELERATORS↗

Modeling Rare Events in High Dimensional Stochastics Systems

The main issue addressed in this note is the study of an algorithm to accelerate the computation of kinetic rates in the context of molecular dynamics(MD). It is based on parallel simulations of short-time trajectories and its main components are: a decomposition of phase space into macrostates or cells, a resampling procedure that ensures that the number of parallel replicas (MD simulations) in each macro-state remains constant, the use of multiple populations (colored replicas) to compute multiple rates (e.g., forward and backward rates) in one simulation. The method leads to enhancing the sampling of macro-states associated to the transition states, since in traditional MD these are likely to be depleted even after short-time simulations. By allowing parallel replicas to carry different probabilistic weights, the number of replicas within each macro-state can be maintained constant without introducing any bias. The empirical variance of the estimated reaction rate, defined as a probability flux, is expectedly diminished. This note is a first attempt towards a better mathematical and numerical understanding of this method. It provides first a mathematical formalization of the notion of colors. Then, the link between this algorithm and a set of closely related methods having been proposed in the literature within the past few years is discussed. Lastly, numerical results are provided that illustrate the efficiency of the method.

59 BASIC BIOLOGICAL SCIENCES↗

Illinois Accelerator Research Center (IARC) [Slides]

There have been and will continue to be new technologies developed in the pursuit of basic science. (Accelerators, Detectors, Magnets, Computing); IARC is focused on developing accelerator based technologies to the point where they are attractive to industry; Through IARC we can leverage Fermi’s human capital, facilities and technologies; We believe that the technologies we are developing at IARC will enable new fields and demonstrate Fermi’s impact, beyond basic discovery science, on the nation’s health, security and wealth.

43 PARTICLE ACCELERATORS↗

Theory and Computation for Mesoscopic Materials Modeling

Accurate and efficient computational modeling of material behavior is essential to the DOE's mission of advancing the development of devices and components needed for power generation and storage. One of the outstanding challenges in atomistic simulation of condensed systems, such as solids, liquids, and glasses, is access to experimentally meaningful length and time scales. The project developed and improved several solution methods to overcome these challenges: diffusive molecular dynamics, adaptive kinetic Monte Carlo, coarse-graining of transition state theory, finite temperature coarse-graining methods for crystalline defects, and optimization-based atomistic-to-continuum coupling methods.

97 MATHEMATICS AND COMPUTING↗