Search NASA⌕ Search

SEARCH · Search NASA

Results for “hardware accelerators”

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 325 records · Page 18

Towards Precision-Aware Fault Tolerance Approaches for Mixed-Precision Applications

Graphics Processing Units (GPUs), the dominantly adopted accelerators in HPC systems, are susceptible to transient hardware fault. New generation of GPUs feature mixed-precision architectures such as NVIDIA Tensor Cores to accelerate matrix multiplications. While being widely adapted, how would they behave under transient hardware faults remain unclear. In this study, we conduct a large-scale fault injection experiments on GEMM kernels implemented with different floating-point data types on the V100 and A100 Tensor Cores, and show distinct error resilience characteristics for the GEMMS with different formats. In the future, we plan to explore this space by building precision-aware floating-point fault tolerance techniques for applications such as DNNs that exercise low-precision computations.

Fang, Bo↗

EMMA: an adaptive mesh refinement cosmological simulation code with radiative transfer

EMMA is a cosmological simulation code aimed at investigating the reionization epoch. It handles simultaneously collisionless and gas dynamics, as well as radiative transfer physics using a moment-based description with the M1 approximation. Field quantities are stored and computed on an adaptive three-dimensional mesh and the spatial resolution can be dynamically modified based on physically motivated criteria. Physical processes can be coupled at all spatial and temporal scales. We also introduce a new and optional approximation to handle radiation: the light is transported at the resolution of the non-refined grid and only once the dynamics has been fully updated, whereas thermo-chemical processes are still tracked on the refined elements. Such an approximation reduces the overheads induced by the treatment of radiation physics. A suite of standard tests are presented and passed by EMMA, providing a validation for its future use in studies of the reionization epoch. The code is parallel and is able to use graphics processing units (GPUs) to accelerate hydrodynamics and radiative transfer calculations. Depending on the optimizations and the compilers used to generate the CPU reference, global GPU acceleration factors between ×3.9 and ×16.9 can be obtained. As a result, vectorization and transfer operations currently prevent better GPU performance and we expect that future optimizations and hardware evolution will lead to greater accelerations.

79 ASTRONOMY AND ASTROPHYSICS↗

Real-Time Target Position Estimation Using Stereoscopic Kilovoltage/Megavoltage Imaging and External Respiratory Monitoring for Dynamic Multileaf Collimator Tracking

Purpose: To develop a real-time target position estimation method using stereoscopic kilovoltage (kV)/megavoltage (MV) imaging and external respiratory monitoring, and to investigate the performance of a dynamic multileaf collimator tracking system using this method. Methods and Materials: The real-time three-dimensional internal target position estimation was established by creating a time-varying correlation model that connected the external respiratory signals with the internal target motion measured intermittently using kV/MV imaging. The method was integrated into a dynamic multileaf collimator tracking system. Tracking experiments were performed for 10 thoracic/abdominal traces. A three-dimensional motion platform carrying a gold marker and a separate one-dimensional motion platform were used to reproduce the target and external respiratory motion, respectively. The target positions were detected by kV (1 Hz) and MV (5.2 Hz) imaging, and external respiratory motion was captured by an optical system (30 Hz). The beam-target alignment error was quantified as the positional difference between the target and circular beam center on the MV images acquired during tracking. The correlation model error was quantified by comparing a model estimate and measured target positions. Results: The root-mean-square errors in the beam-target alignment that had ranged from 3.1 to 7.6 mm without tracking were reduced to <1.5 mm with tracking, except during the model building period (6 s). The root-mean-square error in the correlation model was submillimeters in all directions. Conclusion: A novel real-time target position estimation method was developed and integrated into a dynamic multileaf collimator tracking system and demonstrated an average submillimeter geometric accuracy after initializing the internal/external correlation model. The method used hardware tools available on linear accelerators and therefore shows promise for clinical implementation.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

WE-A-BRF-01: Dual-Energy CT Imaging in Diagnostic Imaging and Radiation Therapy

The quantification accuracy of dual-energy imaging is influenced by the fundamentals of x-ray physics, system geometry, data acquisition hardware/protocol, system calibration, and image processing technique. This symposium will provide updates on the following advanced application areas: Mammography. Volumetric breast density techniques based on standard mammograms require estimation of breast thickness, which is difficult to accurately measure. By comparison, calculation of breast density using dual energy mammography does not require measurement of breast thickness. Dual energy mammography has been implemented using both energy integrating flat panel detectors in conjunction with beam energy switching and energy resolved photon counting detectors. These techniques have been optimized using simulation studies and validated using physical phantoms and postmortem breasts. Chemical decomposition was used as the gold standard for volumetric breast density measurement in postmortem breasts. Breast density measurements have also been compared with results from four-category BI-RADS density rankings, standard image thresholding and Fuzzy k-mean clustering techniques. These studies indicate that dual energy mammography can be used to accurately measure volumetric breast density. Cardiovascular CT. The predicative accuracy of risk models for recurrent stroke and cardiac arrest depends heavily on accurate differentiation of thrombus or calcium from iodine in left atrial appendage or coronary arteries. The amount of energy separation is constrained by image noise; therefore, optimal kVp, beam filtration, and balanced flux are essential for the quantification accuracy of iodine and calcium. The basis materials are combined linearly to generate monochromatic energy images, where CT# accuracy and CNR are energy dependent. With optimal monochromatic energy, the mean iodine concentration for the thrombus, circulatory stasis, and control groups are significantly different. Risk classification based on calcium scores shows excellent agreement with classification on the basis of conventional coronary artery calcium scoring. These studies demonstrate dual-energy cardiovascular CT can potentially be a noninvasive and sensitive modality in high risk patients. On-board KV/MV Imaging. To enhance soft tissue contrast and reduce metal artifacts, we have developed a dual-energy CBCT technique and a novel on-board kV/MV imaging technique based on hardware available on modern linear accelerators. We have also evaluated the feasibility of these two techniques in various phantom studies. Optimal techniques (energy, beam filtration, # of overlapping projections, etc) have been investigated with unique calibration procedures, which leads to successful decomposition of imaged material into acrylic-aluminum basis material pair. This enables the synthesis of virtual monochromatic (VM) CBCT images that demonstrate much less beam hardening, significantly reduced metal artifacts, and/or higher soft tissue CNR compared to single-energy CBCT. Adaptive Radiation Therapy. DECT could actually contribute to the area of Dose-Guided Radiation Therapy (or Adaptive Therapy). The application of DECT imaging using 80kV and 140 kV combinations could potentially increase the image quality by reducing the bone or high density material artifacts and also increase the soft tissue contrast by a light contrast agent. The result of this higher contrast / quality images is beneficial for deformable image registration / segmentation algorithm to improve its accuracy hence to make adaptive therapy less time consuming in its recontouring process. The real time re-planning prior to per treatment fraction could become more realistic with this improvement especially in hypofractional SBRT cases. Learning Objectives: Learn recent developments of dual-energy imaging in diagnosis and radiation therapy; Understand the unique clinical problem and required quantification accuracy in each application; Understand the different approaches to optimize dual-energy imaging techniques for different applications.

60 APPLIED LIFE SCIENCES↗

First proton–proton collisions at the LHC as observed with the ALICE detector: measurement of the charged-particle pseudorapidity density at $\sqrt{s}=900$ GeV

On 23rd November 2009, during the early commissioning of the CERN Large Hadron Collider (LHC), two counter-rotating proton bunches were circulated for the first time concurrently in the machine, at the LHC injection energy of 450 GeV per beam. Although the proton intensity was very low, with only one pilot bunch per beam, and no systematic attempt was made to optimize the collision optics, all LHC experiments reported a number of collision candidates. In the ALICE experiment, the collision region was centred very well in both the longitudinal and transverse directions and 284 events were recorded in coincidence with the two passing proton bunches. The events were immediately reconstructed and analyzed both online and offline. We have used these events to measure the pseudorapidity density of charged primary particles in the central region. In the range |η|<0.5, we obtain d N ch /dη=3. 10±0. 13(stat.)±0.22(syst.) for all inelastic interactions, and d N ch /dη=3.51±0. 15(stat.)±0. 25(syst.) for non-single diffractive interactions. These results are consistent with previous measurements in proton-antiproton interactions at the same centre-of-mass energy at the CERN Sp$\bar{p}$S collider. They also illustrate the excellent functioning and rapid progress of the LHC accelerator, and of both the hardware and software of the ALICE experiment, in this early start-up phase.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Using Likwid and Byfl to Benchmark Hardware Performance

This paper outlines a benchmarking study conducted during my internship at LANL, focusing on CPU (Computer Processing Unit) and program performance assessment. The primary goal was to gather memory access data using three methods across five polybench kernels The data gathered would then be used to compare and contrast to one another and calculate operational intensity for performance comparisons. Benchmarking tools like Byfl and Likwid were employed, with Byfl offering hardware-independent data through LLVM compiler communication and Likwid directly interacting with computer hardware. The study considered various benchmarking factors, including optimization levels, Big O notation ((n)), CPU diversity and specific kernel equations. Big O notation was utilized to simplify code complexity, with detailed breakdwons of operations and memory components for each polybench application. Specific O(n) equations enabled nuanced kernel compariosns, facilitating the identification of performance variations. CPU efficiency assessments were conducted using Likwid tests on two CPUs. The central focus on code optimization aimed at achieving higher speeds and reduced memory usage through streamlined code. Future work propsoes creating a roofline model, synthesizing benchmarking data into a comprehensive data graph to assist in optimizing code and improving hardware performance. The potential impact on the laboratory or national mission was underscored, emphasizing the importance of optimizing applications and hardware to conserve resources and accelerate program execution. The specific relevance to LANL’s operations in math-intensive fields such as Nuclear Fission, Space Exploration, and Nanotechnology highlights the necessity of efficient benchmarking for resource conservation and proram speed. Overall, this study contributes to the understanding of CPU and program performance, providing insights for future optimization efforts in a laboratory setting

97 MATHEMATICS AND COMPUTING↗

Final Seismic Shake Table Test Plan

The Spent Fuel Waste Disposition (SFWD) program is planning to conduct a full-scale seismic shake table test on the dry storage systems of spent nuclear fuel (SNF) to close the gap related to seismic loads on fuel assemblies in dry storage systems. This test will allow for quantifying the strains and accelerations on surrogate fuel assembly hardware and cladding during earthquakes of different magnitudes and frequency content. Full-scale testing is needed because a dry storage system is a complex and highly nonlinear system making it hard to predict (model) the responses to seismic excitations. The non-linearity arises from the multiple spatial gaps in the system – between fuel rods and the basket, between the basket and dry storage canister, between the dry storage canister and the storage cask (overpack), and ventilation gaps. The non-linearities pose significant limitations on the value of tests with scaled systems.

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗

Particle Impact Simulation and Ignition Prediction

An experimentally calibrated tool is needed to predict if a system is susceptible to failure by particle impact ignition (PI) based on use conditions, materials, and flow geometry. This tool will accelerate new components, evaluating existing hardware, and help disposition anomalies. Conduct particle impact testing with in-situ diagnostics and complementary simulations on subset of key engineering materials (IN718, M400, 316L, 6061, Ti64, Zr) to develop a proof-of-concept predictive tool for assessing the risk of PI for idealized geometries (spherical particles) in realistic environments. Assess particle/target interactions (coefficient of restitution, ignition, kindling) using instrumented particle impact rigs while systematically varying key parameters (materials, particle size, environment, target configuration). Determine key field variables (temperature, strain, stress) in particle impacts using Multiphysics finite element and hydrocode simulations validated through comparison with experimental measurements and observations. Synthesize experiments and simulations into constitutive models for PI that can be integrated with existing computational fluid dynamics (CFD) and Debris Transport Analysis (DTA) tools in future efforts

particle impact↗

Particle Impact Simulation and Ignition Prediction

An experimentally calibrated tool is needed to predict if a system is susceptible to failure by particle impact ignition (PI) based on use conditions, materials, and flow geometry. This tool will accelerate new components, evaluating existing hardware, and help disposition anomalies. - Conduct particle impact testing with in-situ diagnostics and complementary simulations on subset of key engineering materials (IN718, M400, 316L, 6061, Ti64, Zr) to develop a proof-of-concept predictive tool for assessing the risk of PI for idealized geometries (spherical particles) in realistic environments. - Assess particle/target interactions (coefficient of restitution, ignition, kindling) using instrumented particle impact rigs while systematically varying key parameters (materials, particle size, environment, target configuration). - Determine key field variables (temperature, strain, stress) in particle impacts using Multiphysics finite element and hydrocode simulations validated through comparison with experimental measurements and observations. - Synthesize experiments and simulations into constitutive models for PI that can be integrated with existing computational fluid dynamics (CFD) and Debris Transport Analysis (DTA) tools in future efforts.

Jonathan Tylka↗

Acceleration of the matrix multiplication of Radiance three phase daylighting simulations with parallel computing on heterogeneous hardware of personal computer

Building designers are increasingly relying on complex fenestration systems to reduce energy consumed for lighting and HVAC in low energy buildings. Radiance, a lighting simulation program, has been used to conduct daylighting simulations for complex fenestration systems. Depending on the configurations, the simulation can take hours or even days using a personal computer. This paper describes how to accelerate the matrix multiplication portion of a Radiance three-phase daylight simulation by conducting parallel computing on heterogeneous hardware of a personal computer. The algorithm was optimized and the computational part was implemented in parallel using OpenCL. The speed of new approach was evaluated using various daylighting simulation cases on a multicore central processing unit and a graphics processing unit. Based on the measurements and analysis of the time usage for the Radiance daylighting simulation, further speedups can be achieved by using fast I/O devices and storing the data in a binary format.

29 ENERGY PLANNING, POLICY, AND ECONOMY↗

Kernel fusion in atomistic spin dynamics simulations on Nvidia GPUs using tensor core

In atomistic spin dynamics simulations, the time cost of constructing the space- and time-displaced pair correlation function in real space increases quadratically as the number of spins N, leading to significant computational effort. The GEMM subroutine can be adopted to accelerate the calculation of the dynamical spin-spin correlation function, but the computational cost of simulating large spin systems (>40000 spins) on CPUs remains expensive. In this work, we perform the simulation on the graphics processing unit (GPU), a hardware solution widely used as an accelerator for scientific computing and deep learning. Here we show that GPUs can accelerate the simulation up to 25-fold compared to multi-core CPUs when using the GEMM subroutine on both. To hide memory latency, we fuse the element-wise operation into the GEMM kernel using CUTLASS that can improve the performance by 26% ~ 33% compared to implementation based on cuBLAS. Furthermore, we perform the on-the-fly calculation in the epilogue of the GEMM subroutine to avoid saving intermediate results on global memory, which makes the large-scale atomistic spin dynamics simulation feasible and affordable.

97 MATHEMATICS AND COMPUTING↗

Basic features of the STS/Spacelab vibration environment

The Space Shuttle acceleration environment is characterized. The acceleration environment is composed of a residual or quasi-steady component and higher frequency components induced by vehicle structural modes and the operation of onboard machinery. Quasi-steady accelerations are generally due to atmospheric drag, gravity gradient effects, and rotational forces. These accelerations tend to vary with the orbital frequency (approx. 10(exp -4) Hz) and have magnitudes less than or equal to 10(exp -6) g(sub 0) (where 1 g(sub 0) is terrestrial gravity). Higher frequency g-jitter is characterized by oscillatory disturbances in the 1-100 Hz range and transient components. Oscillatory accelerations are related to the response of large flexible structures like antennae, the Spacelab module, and the Orbiter itself, and to the operation of rotating machinery. The Orbiter structural modes in the 1-10 Hz range, are excited by oscillatory and transient disturbances and tend to dominate the energy spectrum of the acceleration environment. A comparison of the acceleration measurements from different Space Shuttle missions reveals the characteristic signature of the structural modes of the Orbiter overlaid with mission specific hardware induced disturbances and their harmonics. Transient accelerations are usually attributed to crew activity and Orbiter thruster operations. During crew sleep periods, the acceleration levels are typically on the order of 10(exp -6) g(sub 0) (1 micro-g). Crew work and exercise tend to raise the accelerations to the 10(exp -3) g(sub 0) (1 milli-g) level. Vernier reaction control system firings tend to cause accelerations of 10(exp -4) g(sub 0), while primary reaction control system and Orbiter maneuvering system firings cause accelerations as large as 10(exp -2) g(sub 0). Vibration isolation techniques (both active and passive systems) used during crew exercise have been shown to significantly reduce the acceleration magnitudes.

Baugher, Charles R.↗

Multichannel Analysis of Surface Waves Accelerated (MASWAccelerated): Software for efficient surface wave inversion using MPI and GPUs

Multichannel Analysis of Surface Waves (MASW) is a technique frequently used in geotechnical engineering and engineering geophysics to infer 1D layered models of seismic shear wave velocities in the top tens to hundreds of meters of the subsurface. We aim to accelerate MASW calculations by capitalizing on modern computer hardware available in the workstations of most engineers: multiple cores and graphics processing units (GPUs). We propose new parallel and GPU accelerated algorithms for computing 1D MASW inversion, and provide software implementations in C using Message Passing Interface (MPI) and CUDA. These algorithms take advantage of sparsity that arises in the problem, and the work balance between processes considers typical data trends. We compare our methods to an existing open source Matlab MASW tool. Our serial C implementation achieves a 2x speedup over the Matlab software, and we continue to see improvements by parallelizing the problem with MPI. Here we see nearly perfect strong and weak scaling for uniform data, and improve strong scaling for realistic data by repartitioning the problem to process mapping. By utilizing GPUs available on most modern workstations, we observe an additional 1.3x speedup over the serial C implementation on the first use of the method. We typically repeatedly evaluate theoretical dispersion curves as part of an optimization procedure, and on the GPU the kernel can be cached for faster reuse on later runs. We observe a 3.2x speedup on the cached GPU runs compared to the serial C runs. This work is the first open-source parallel or GPU-accelerated software tool for MASW imaging, and should enable geotechnical engineers to fully utilize all computer hardware at their disposal.

58 GEOSCIENCES↗

PREPARING FOR EXASCALE: ORNL Leadership Computing Application Requirements and Strategy

In 2009 the Oak Ridge Leadership Computing Facility (OLCF), a U.S. Department of Energy (DOE) facility at the Oak Ridge National Laboratory (ORNL) National Center for Computational Sciences (NCCS), elicited petascale computational science requirements from leading computational scientists in the international science community. This effort targeted science teams whose projects received large computer allocation awards on OLCF systems. A clear finding of this process was that in order to reach their science goals over the next several years, multiple projects will require computational resources in excess of an order of magnitude more powerful than those currently available. Additionally, for the longer term, next-generation science will require computing platforms of exascale capability in order to reach DOE science objectives over the next decade. It is generally recognized that achieving exascale in the proposed time frame will require disruptive changes in computer hardware and software. Processor hardware will become necessarily heterogeneous and will include accelerator technologies. Software must undergo the concomitant changes needed to extract the available performance from this heterogeneous hardware. This disruption portends to be substantial, not unlike the change to the message passing paradigm in the computational science community over 20 years ago. Since technological disruptions take time to assimilate, we must aggressively embark on this course of change now, to insure that science applications and their underlying programming models are mature and ready when exascale computing arrives. This includes initiation of application readiness efforts to adapt existing codes to heterogeneous architectures, support of relevant software tools, and procurement of next-generation hardware testbeds for porting and testing codes. The 2009 OLCF requirements process identified numerous actions necessary to meet this challenge: (1) Hardware capabilities must be advanced on multiple fronts, including peak flops, node memory capacity, interconnect latency, interconnect bandwidth, and memory bandwidth. (2) Effective parallel programming interfaces must be developed to exploit the power of emerging hardware. (3) Science application teams must now begin to adapt and reformulate application codes to the new hardware and software, typified by hierarchical and disparate layers of compute, memory and concurrency. (4) Algorithm research must be realigned to exploit this hierarchy. (5) When possible, mathematical libraries must be used to encapsulate the required operations in an efficient and useful way. (6) Software tools must be developed to make the new hardware more usable. (7) Science application software must be improved to cope with the increasing complexity of computing systems. (8) Data management efforts must be readied for the larger quantities of data generated by larger, more accurate science models. Requirements elicitation, analysis, validation, and management comprise a difficult and inexact process, particularly in periods of technological change. Nonetheless, the OLCF requirements modeling process is becoming increasingly quantitative and actionable, as the process becomes more developed and mature, and the process this year has identified clear and concrete steps to be taken. This report discloses (1) the fundamental science case driving the need for the next generation of computer hardware, (2) application usage trends that illustrate the science need, (3) application performance characteristics that drive the need for increased hardware capabilities, (4) resource and process requirements that make the development and deployment of science applications on next-generation hardware successful, and (5) summary recommendations for the required next steps within the computer and computational science communities.

applications requirements↗

Diagnostic emulation: Implementation and user's guide

The Diagnostic Emulation Technique was developed within the System Validation Methods Branch as a part of the development of methods for the analysis of the reliability of highly reliable, fault tolerant digital avionics systems. This is a general technique which allows for the emulation of a digital hardware system. The technique is general in the sense that it is completely independent of the particular target hardware which is being emulated. Parts of the system are described and emulated at the logic or gate level, while other parts of the system are described and emulated at the functional level. This algorithm allows for the insertion of faults into the system, and for the observation of the response of the system to these faults. This allows for controlled and accelerated testing of system reaction to hardware failures in the target machine. This document describes in detail how the algorithm was implemented at NASA Langley Research Center and gives instructions for using the system.

Becher, Bernice↗

EXPERIMENTAL STUDIES OF NONLINEAR INTEGRABLE OPTICS

State-of-the-art accelerators at energy and intensity frontiers require increasingly bright and powerful particle beams. In conventional linear lattices, intense beams suffer from collective instabilities, resulting in beam losses and maximum beam intensity limits. This thesis presents experimental studies of a novel lattice design concept, the nonlinear integrable optics (NIO), aimed at enhancing beam stability limits with little to no beam dynamics degradation. Single-particle beam dynamics measurements of two NIO devices, the quasi-integrable octupole system and the fully integrable Danilov-Nagaitsev system, were carried out at the purpose-built Integrable Optics Test Accelerator (IOTA) at Fermilab. Their simulation, hardware design, and commissioning process are presented. Extensive model and analysis algorithm development and benchmarking is described. Electron beam data from two scientific runs is analyzed, yielding frequency and phase space dynamics consisten t with m odels. These results demonstrate viability and advantages of the NIO design, providing the groundwork for proton studies in the strong space-charge regime and future integrable accelerators.

Kuklev, Nikita↗

FPGA-based HPC accelerators: An evaluation on performance and energy efficiency

Hardware specialization is a promising direction for the future of digital computing. Reconfigurable technologies enable hardware specialization with modest non-recurring engineering cost, but their performance and energy efficiency compared to state-of-the-art processor architectures remain an open question. In this article, we use FPGAs to evaluate the benefits of building specialized hardware for numerical kernels found in scientific applications. In order to properly evaluate performance, we not only compare Intel Arria 10 and Xilinx U280 performance against Intel Xeon, Intel Xeon Phi, and NVIDIA V100 GPUs, but we also extend the Empirical Roofline Toolkit (ERT) to FPGAs in order to assess our results in terms of the Roofline model. We show design optimization and tuning techniques for peak FPGA performance at reasonable hardware usage and power consumption. As FPGA peak performance is known to be far less than that of a GPU, we also benchmark the energy efficiency of each platform for the scientific kernels comparing against microbenchmark and technological limits. Results show that while FPGAs struggle to compete in absolute terms with GPUs on memory- and compute-intensive kernels, they require far less power and can deliver nearly the same energy efficiency.

97 MATHEMATICS AND COMPUTING↗