Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “limited memory”

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 523 records · Page 29
All sourcesNASA NTRSDOE OSTI
DOE OSTI2023

Nonvolatile Memristive Effect in Few-Layer CrI 3 Driven by Electrostatic Gating

The potential of memristive devices for applications in nonvolatile memory and neuromorphic computing has sparked considerable interest, particularly in exploring memristive effects in two-dimensional (2D) magnetic materials. However, the progress in developing nonvolatile, magnetic field-free memristive devices using 2D magnets has been limited. In this work, we report an electrostatic-gating-induced nonvolatile memristive effect in CrI 3 -based tunnel junctions. The few-layer CrI 3 -based tunnel junction manifests notable hysteresis in its tunneling resistance as a function of gate voltage. We further engineered a nonvolatile memristor using the CrI 3 tunneling junction with low writing power and at zero magnetic field. We show that the hysteretic transport observed is not a result of trivial effects or inherent magnetic properties of CrI 3 . We propose a potential association between the memristive effect and the newly predicted ferroelectricity in CrI 3 via gating-induced Jahn–Teller distortion. Furthermore, our work illuminates the potential of 2D magnets in developing next-generation advanced computing technologies.

2D magnet↗
DOE OSTI2023

Ultra-low latency recurrent neural network inference on FPGAs for physics applications with hls4ml

Abstract Recurrent neural networks have been shown to be effective architectures for many tasks in high energy physics, and thus have been widely adopted. Their use in low-latency environments has, however, been limited as a result of the difficulties of implementing recurrent architectures on field-programmable gate arrays (FPGAs). In this paper we present an implementation of two types of recurrent neural network layers—long short-term memory and gated recurrent unit—within the hls4ml framework. We demonstrate that our implementation is capable of producing effective designs for both small and large models, and can be customized to meet specific design requirements for inference latencies and FPGA resources. We show the performance and synthesized designs for multiple neural networks, many of which are trained specifically for jet identification tasks at the CERN Large Hadron Collider.

97 MATHEMATICS AND COMPUTING↗
NASA NTRS1982

Computational fluid dynamics - The coming revolution

The development of aerodynamic theory is traced from the days of Aristotle to the present, with the next stage in computational fluid dynamics dependent on superspeed computers for flow calculations. Additional attention is given to the history of numerical methods inherent in writing computer codes applicable to viscous and inviscid analyses for complex configurations. The advent of the superconducting Josephson junction is noted to place configurational demands on computer design to avoid limitations imposed by the speed of light, and a Japanese projection of a computer capable of several hundred billion operations/sec is mentioned. The NASA Numerical Aerodynamic Simulator is described, showing capabilities of a billion operations/sec with a memory of 240 million words using existing technology. Near-term advances in fluid dynamics are discussed.

Graves, R. A., Jr.↗
NASA NTRS1984

IMAGES: A digital computer program for interactive modal analysis and gain estimation for eigensystem synthesis

An interactive digital computer program for modal analysis and gain estimation for eigensystem synthesis was written. Both mathematical and operation considerations are described; however, the mathematical presentation is limited to those concepts essential to the operational capability of the program. The program is capable of both modal and spectral synthesis of multi-input control systems. It is user friendly, has scratchpad capability and dynamic memory, and can be used to design either state or output feedback systems.

Jones, R. L.↗
NASA NTRS1991

Memory implications of the spin-value effect in soft multilayers

New multilayer permalloy materials with thin separation layers that have significantly increased M-R coefficients at room temperature have the potential for substantially increasing both the speed and density of M-R memories. An analysis in the following discussion shows that the bit density at a given performance level is related to the product of M-R response and sheet resistivity when operating at the current-density limit. With a doubling of both sheet resistivity and M-R response over existing materials, the bit area can be reduced by about a factor of 4. Alternatively, if the bit area is kept the same, signal output is increased by a factor of 4 and circuit time constants can be reduced by a factor of 16.

Pohm, A. V.↗
NASA NTRS2004

Finite Element Analysis in Concurrent Processing: Computational Issues

The purpose of this research is to investigate the potential application of new methods for solving large-scale static structural problems on concurrent computers. It is well known that traditional single-processor computational speed will be limited by inherent physical limits. The only path to achieve higher computational speeds lies through concurrent processing. Traditional factorization solution methods for sparse matrices are ill suited for concurrent processing because the null entries get filled, leading to high communication and memory requirements. The research reported herein investigates alternatives to factorization that promise a greater potential to achieve high concurrent computing efficiency. Two methods, and their variants, based on direct energy minimization are studied: a) minimization of the strain energy using the displacement method formulation; b) constrained minimization of the complementary strain energy using the force method formulation. Initial results indicated that in the context of the direct energy minimization the displacement formulation experienced convergence and accuracy difficulties while the force formulation showed promising potential.

Sobieszczanski-Sobieski, Jaroslaw↗
DOE OSTI2021

Higher-order particle representation for particle-in-cell simulations

In this paper we present an alternative approach to the representation of simulation particles for unstructured electrostatic and electromagnetic PIC simulations. In our modified PIC algorithm we represent particles as having a smooth shape function limited by some specified finite radius, r 0 . A unique feature of our approach is the representation of this shape by surrounding simulation particles with a set of virtual particles with delta shape, with fixed offsets and weights derived from Gaussian quadrature rules and the value of r 0 . As the virtual particles are purely computational, they provide the additional benefit of increasing the arithmetic intensity of traditionally memory bound particle kernels. The modified algorithm is implemented within Sandia National Laboratories' unstructured EMPIRE-PIC code, for electrostatic and electromagnetic simulations, using periodic boundary conditions. We show results for a representative set of benchmark problems, including electron orbit, a transverse electromagnetic wave propagating through a plasma, numerical heating, and a plasma slab expansion. In this work, good error reduction across all of the chosen problems is achieved as the particles are made progressively smoother, with the optimal particle radius appearing to be problem-dependent.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
NASA NTRS1983

High-speed digital signal normalization for feature identification

A design approach for high speed normalization of digital signals was developed. A reciprocal look up table technique is employed, where a digital value is mapped to its reciprocal via a high speed memory. This reciprocal is then multiplied with an input signal to obtain the normalized result. Normalization improves considerably the accuracy of certain feature identification algorithms. By using the concept of pipelining the multispectral sensor data processing rate is limited only by the speed of the multiplier. The breadboard system was found to operate at an execution rate of five million normalizations per second. This design features high precision, a reduced hardware complexity, high flexibility, and expandability which are very important considerations for spaceborne applications. It also accomplishes a high speed normalization rate essential for real time data processing.

Ortiz, J. A.↗
NASA NTRS2010

Remembrance of Things Future: Prospective Memory in Laboratory, Workplace, and Everyday Settings

In this review, oriented to the human factors community, I will summarize and provide a perspective on recent research and theory on prospective memory. This will not be an exhaustive review of literature, which is already available in two excellent recent books that provide a wealth of detail on the current state of experimental research (Kliegel, McCaniel, & Einstein, 2008; McDaniel & Einstein, 2007; also see Brandimonte, Einstein, & McDaniel, 1996, for a still relevant overview of the field as it was emerging). Rather, I will explore the limits of existing experimental paradigms and theory, Vvilich, in my opinion, fail to capture some critical aspects of performance outside the laboratory. I will also review the relatively few studies in workplace and everyday settings and will discuss several studies that attempt to bridge between the bulk of experimental studies and these few naturalistic studies. Finally, I will describe countermeasures that can reduce vulnerability to forgetting to perform intended tasks, and I will propose a research agenda that would extend existing experimental and theoretical approaches and would support human factors practitioners by generating information on a wide range of issues relevant to prospective memory performance in natural settings.

Dismukes, R. Key↗
DOE OSTI2020

GPU-acceleration of the ELPA2 distributed eigensolver for dense symmetric and hermitian eigenproblems

The solution of eigenproblems is often a key computational bottleneck that limits the tractable system size of numerical algorithms, among them electronic structure theory in chemistry and in condensed matter physics. Large eigenproblems can easily exceed the capacity of a single compute node, thus must be solved on distributed-memory parallel computers. We here present GPU-oriented optimizations of the ELPA two-stage tridiagonalization eigensolver (ELPA2). On top of cuBLAS-based GPU offloading, we add a CUDA kernel to speed up the back-transformation of eigenvectors, which can be the computationally most expensive part of the two-stage tridiagonalization algorithm. Furthermore, we benchmark the performance of this GPU-accelerated eigensolver on two hybrid CPU–GPU architectures, namely a compute cluster based on Intel Xeon Gold CPUs and NVIDIA Volta GPUs, and the Summit supercomputer based on IBM POWER9 CPUs and NVIDIA Volta GPUs. Consistent with previous benchmarks on CPU-only architectures, the GPU-accelerated two-stage solver exhibits a parallel performance superior to the one-stage counterpart. Finally, we demonstrate the performance of the GPU-accelerated eigensolver developed in this work for routine semi-local KS-DFT calculations comprising thousands of atoms.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2023

XploreNAS : Explore Adversarially Robust and Hardware-efficient Neural Architectures for Non-ideal Xbars

Compute In-Memory platforms such as memristive crossbars are gaining focus as they facilitate acceleration of Deep Neural Networks (DNNs) with high area and compute efficiencies. However, the intrinsic non-idealities associated with the analog nature of computing in crossbars limits the performance of the deployed DNNs. Furthermore, DNNs are shown to be vulnerable to adversarial attacks leading to severe security threats in their large-scale deployment. Thus, finding adversarially robust DNN architectures for non-ideal crossbars is critical to the safe and secure deployment of DNNs on the edge. This work proposes a two-phase algorithm-hardware co-optimization approach called XploreNAS that searches for hardware efficient and adversarially robust neural architectures for non-ideal crossbar platforms. We use the one-shot Neural Architecture Search approach to train a large Supernet with crossbar-awareness and sample adversarially robust Subnets therefrom, maintaining competitive hardware efficiency. Our experiments on crossbars with benchmark datasets (SVHN, CIFAR10, CIFAR100) show up to ~8–16% improvement in the adversarial robustness of the searched Subnets against a baseline ResNet-18 model subjected to crossbar-aware adversarial training. We benchmark our robust Subnets for Energy-Delay-Area-Products (EDAPs) using the Neurosim tool and find that with additional hardware efficiency–driven optimizations, the Subnets attain ~1.5–1.6× lower EDAPs than ResNet-18 baseline.

97 MATHEMATICS AND COMPUTING↗
NASA NTRS2000

All-Metal Magnetic RAM

The factors that enter into the development of an all-metal, nonvolatile magnetic RAM, in which multilayer giant magnetoresistive films are used for all functions - storage, readout, and support electronics - are described. Four significant characteristics are expected to favor all-metal over hybrid magnetic RAM. First, silicon-technology fabrication requires a large number of masking steps, including complex ones such as ion implantation. Conversely, all-metal technology is inherently simple: fewer masking steps, no doping, scaling to lithographic limits, very little operating power. Second, the all-metal footprint is significantly smaller than the hybrid one. Third, an all-metal RAM is expected to be able to be miniaturized to lithographic limits; miniaturization of hybrid magnetic RAMs is likely to be limited by the semiconductor circuitry. Finally, semiconductor processing and magnetic processing in MRAM are done separately because the former requires high temperatures, whereas magnetic fabrication is a low-temperature process. By contrast, because both GMR electronics and the memory elements are made of the same materials, the two major components are deposited and patterned concurrently on the same substrate.

Torok, E. J.↗
NASA NTRS2003

2003 NASA Faculty Fellowship Program at Glenn Research Center

The Office of Education at NASA Headquarters provides overall policy and direction for the NASA Faculty Fellowship Program (NFFP). The American Society for Engineering Education (ASEE) and the Universities Space Research Association (USRA) have joined in partnership to recruit participants, accept applications from a broad range of participants, and provide overall evaluation of the NFFP. The NASA Centers, through their University Affairs Officers, develop and operate the experiential part of the program. In concert with co-directing universities and the Centers, Fellows are selected and provided the actual research experiences. This report summarizes the 2003 session conducted at the Glenn Research Center (GRC).Research topics covered a variety of areas including, but not limited to, biological sensors, modeling of biological fluid systems, electronic circuits, ceramics and coatings, unsteady probablistic analysis and aerodynamics, gas turbines, environmental monitoring systems for water quality, air quality, gaseous and particulate emissions, bearings for flywheel energy storage, shape memory alloys,photonic interrogation and nanoprocesses,carbon nanotubes, polymer synthesis for fuel cells, aviation communications, algorithm development and RESPlan Database.

Prahl, Joseph M.↗
DOE OSTI2025

Nitrogen Vacancies Induce Fatigue in Ferroelectric Al 0.93 B 0.07 N

Wurtzite ferroelectrics (e.g., Al 0.93 B 0.07 N) are being explored for high-temperature and emerging near-, or in-compute, memory architectures due to the material advantages offered by their large remanent polarization and robust chemical stability. Despite these advantages, current Al0.93B0.07N devices do not have sufficient endurance lifetime to meet roadmap targets. To identify the defects responsible for this limited endurance, a combination of electronic measurements and optical spectroscopies characterized the evolution of defect states within Al 0.93 B 0.07 N with cycling. Ultrathin (∼10 nm) metal contacts were used to optically probe regions subject to ferroelectric switching; photoluminescence spectroscopy identified the emergence of a transition near 2.1 eV whose intensity scaled with the non-switching polarization quantified via positive-up negative-down (PUND) measurements. Accompanying thermally stimulated depolarization current (TSDC) and modulus spectroscopy measurements also observed the strengthening of a state near 2.1 eV. The origin of this feature is ascribed to transitions between a nitrogen vacancy and another defect deeper in the bandgap. Recognizing that the impurity concentration is largely fixed, strengthening of this transition indicates an increase in the number of nitrogen vacancies. Switching, therefore, creates vacancies in Al 0.93 B 0.07 N likely due to hot-atom damage induced by the aggressive fields necessary to switch wurtzite materials that ultimately limits endurance.

36 MATERIALS SCIENCE↗
NASA NTRS1972

Job-mix modeling and system analysis of an aerospace multiprocessor.

An aerospace guidance computer organization, consisting of multiple processors and memory units attached to a central time-multiplexed data bus, is described. A job mix for this type of computer is obtained by analysis of Apollo mission programs. Multiprocessor performance is then analyzed using: 1) queuing theory, under certain 'limiting case' assumptions; 2) Markov process methods; and 3) system simulation. Results of the analyses indicate: 1) Markov process analysis is a useful and efficient predictor of simulation results; 2) efficient job execution is not seriously impaired even when the system is so overloaded that new jobs are inordinately delayed in starting; 3) job scheduling is significant in determining system performance; and 4) a system having many slow processors may or may not perform better than a system of equal power having few fast processors, but will not perform significantly worse.

Mallach, E. G.↗
NASA NTRS1979

LED instrument approach instruction display

A display employing light emitting diodes (LED's) was developed to demonstrate the feasibility of such displays for presenting landing and navigation information to reduce the workload of general aviation pilots during IFR flight. The display consists of a paper tape reader, digital memory, control electronics, digital latches, and LED alphanumeric displays. A presentable digital countdown clock-timer is included as part of the system to provide a convenient means of monitoring time intervals for precise flight navigation. The system is a limited capability prototype assembled to test pilot reaction to such a device under simulated IFR operation. Pilot opinion indicates that the display is helpful in reducing the IFR pilots workload when used with a runway approach plate. However, the development of a compact, low power second generation display was recommended which could present several instructions simultaneously and provide information update capability. A microprocessor-based display could fulfill these requirements.

Meredith, B. D.↗
NASA NTRS1982

Evaluation of the SPAR thermal analyzer on the CYBER-203 computer

The use of the CYBER 203 vector computer for thermal analysis is investigated. Strengths of the CYBER 203 include the ability to perform, in vector mode using a 64 bit word, 50 million floating point operations per second (MFLOPS) for addition and subtraction, 25 MFLOPS for multiplication and 12.5 MFLOPS for division. The speed of scalar operation is comparable to that of a CDC 7600 and is some 2 to 3 times faster than Langley's CYBER 175s. The CYBER 203 has 1,048,576 64-bit words of real memory with an 80 nanosecond (nsec) access time. Memory is bit addressable and provides single error correction, double error detection (SECDED) capability. The virtual memory capability handles data in either 512 or 65,536 word pages. The machine has 256 registers with a 40 nsec access time. The weaknesses of the CYBER 203 include the amount of vector operation overhead and some data storage limitations. In vector operations there is a considerable amount of time before a single result is produced so that vector calculation speed is slower than scalar operation for short vectors.

Robinson, J. C.↗
NASA NTRS1993

Efficient partitioning and assignment on programs for multiprocessor execution

The general problem studied is that of segmenting or partitioning programs for distribution across a multiprocessor system. Efficient partitioning and the assignment of program elements are of great importance since the time consumed in this overhead activity may easily dominate the computation, effectively eliminating any gains made by the use of the parallelism. In this study, the partitioning of sequentially structured programs (written in FORTRAN) is evaluated. Heuristics, developed for similar applications are examined. Finally, a model for queueing networks with finite queues is developed which may be used to analyze multiprocessor system architectures with a shared memory approach to the problem of partitioning. The properties of sequentially written programs form obstacles to large scale (at the procedure or subroutine level) parallelization. Data dependencies of even the minutest nature, reflecting the sequential development of the program, severely limit parallelism. The design of heuristic algorithms is tied to the experience gained in the parallel splitting. Parallelism obtained through the physical separation of data has seen some success, especially at the data element level. Data parallelism on a grander scale requires models that accurately reflect the effects of blocking caused by finite queues. A model for the approximation of the performance of finite queueing networks is developed. This model makes use of the decomposition approach combined with the efficiency of product form solutions.

Standley, Hilda M.↗