Search NASA⌕ Search

SEARCH · Search NASA

Results for “memory effects”

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 919 records · Page 51
All sourcesNASA NTRSDOE OSTI
DOE OSTI2014

Voltage induced magnetostrictive switching of nanomagnets: Strain assisted strain transfer torque random access memory

A spintronic device, called the “strain assisted spin transfer torque (STT) random access memory (RAM),” is proposed by combining the magnetostriction effect and the spin transfer torque effect which can result in a dramatic improvement in the energy dissipation relative to a conventional STT-RAM. Magnetization switching in the device which is a piezoelectric-ferromagnetic heterostructure via the combined magnetostriction and STT effect is simulated by solving the Landau-Lifshitz-Gilbert equation incorporating the influence of thermal noise. The simulations show that, in such a device, each of these two mechanisms (magnetostriction and spin transfer torque) provides in a 90° rotation of the magnetization leading a deterministic 180° switching with a critical current significantly smaller than that required for spin torque alone. Such a scheme is an attractive option for writing magnetic RAM cells.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗
DOE OSTI2014

Performance model-directed data sieving for high-performance I/O

Many scientific computing applications and engineering simulations exhibit noncontiguous I/O access patterns. Data sieving is an important technique to improve the performance of noncontiguous I/O accesses by combining small and noncontiguous requests into a large and contiguous request. It has been proven effective even though more data are potentially accessed than demanded. In this study, we propose a new data sieving approach namely performance model-directed data sieving, or PMD data sieving in short. It improves the existing data sieving approach from two aspects: (1) dynamically determines when it is beneficial to perform data sieving; and (2) dynamically determines how to perform data sieving if beneficial. It improves the performance of the existing data sieving approach considerably and reduces the memory consumption as verified by both theoretical analysis and experimental results. Given the importance of supporting noncontiguous accesses effectively and reducing the memory pressure in a large-scale system, the proposed PMD data sieving approach in this research holds a great promise and will have an impact on high-performance I/O systems.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2018

A Heterogeneity-Aware Task Scheduler for Spark

Big data processing systems such as Spark are employed in an increasing number of diverse applications-such as machine learning, graph computation, and scientific computing-each with dynamic and different resource needs. These applications increasingly run on heterogeneous hardware, e.g., with out-of-core accelerators. However, big data platforms do not factor in the multi-dimensional heterogeneity of applications and hardware. This leads to a fundamental mismatch between the application and hardware characteristics, and the resource scheduling adopted in big data platforms. For example, Hadoop and Spark consider only data locality when assigning tasks to nodes, and typically disregard the hardware capabilities and suitability to specific application requirements. In this paper, we present RUPAM, a heterogeneity-aware task scheduling system for big data platforms, which considers both task-level resource characteristics and underlying hardware characteristics, as well as preserves data locality. RUPAM adopts a simple yet effective heuristic to decide the dominant scheduling factor (e.g., CPU, memory, or I/O), given a task in a particular stage. Our experiments show that RUPAM is able to improve the performance of representative applications by up to 62.3% compared to the standard Spark scheduler.

Computer Science↗
NASA NTRS1991

New dynamic FET logic and serial memory circuits for VLSI GaAs technology

The complexity of GaAs field effect transistor (FET) very large scale integration (VLSI) circuits is limited by the maximum power dissipation while the uniformity of the device parameters determines the functional yield. In this work, digital GaAs FET circuits are presented that eliminate the DC power dissipation and reduce the area to 50% of that of the conventional static circuits. Its larger tolerance to device parameter variations results in higher functional yield.

Eldin, A. G.↗
NASA NTRS2008

Radiation effects assessment of MRAM devices

The Magneto-resistive Random Access Memory (MRAM) from Freescale Semiconductor, MR2A16A, was subjected to heavy ion single event testing. Test results show that this device is sensitive to Single Event Latchup (SEL). The sensitivity of the MRAM was attributed to the complementary metal oxide semiconductor (CMOS) process in which the active portion of this device is constructed. Therefore, the device must be used with caution and may require mitigation techniques if used in a space environment. There was no indication that the MRAM technology itself, the memory element construction, is subject to damage from heavy ions.

McClure, Steve↗
DOE OSTI2016

Rosiglitazone attenuates inflammation and CA3 neuronal loss following traumatic brain injury in rats

Rosiglitazone, a potent peroxisome proliferator-activated receptor (PPAR)-γ agonist, has been shown to confer neuroprotective effects in stroke and spinal cord injury, but its role in the traumatic brain injury (TBI) is still controversial. Using a controlled cortical impact model in rats, the current study was designed to determine the effects of rosiglitazone treatment (6 mg/kg at 5 min, 6 h and 24 h post injury) upon inflammation and histological outcome at 21 d after TBI. In addition, the effects of rosiglitazone upon inflammatory cytokine transcription, vestibulomotor behavior and spatial memory function were determined at earlier time points (24 h, 1–5 d, 14–20 d post injury, respectively). Compared with the vehicle-treated group, rosiglitazone treatment suppressed production of TNFα at 24 h after TBI, attenuated activation of microglia/macrophages and increased survival of CA3 neurons but had no effect on lesion volume at 21 d after TBI. Rosiglitazone-treated animals had improved performance on beam balance testing, but there was no difference in spatial memory function as determined by Morris water maze. In summary, this study indicates that rosiglitazone treatment in the first 24 h after TBI has limited anti-inflammatory and neuroprotective effects in rat traumatic injury. Further study using an alternative dosage paradigm and more sensitive behavioral testing may be warranted. - Highlights: • Effects of rosiglitazone after CCI were evaluated using a rat TBI model. • Rosiglitazone suppressed production of TNFα at 24 h after CCI. • Rosiglitazone inhibited microglial activation at 21 d after CCI. • Rosiglitazone increased survival of CA3 neurons at 21 d after CCI. • Rosiglitazone-treated animals had improved performance in beam balance testing.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2023

Abnormal grain growth of 68Cu–16Al–16Zn alloys for elastocaloric cooling via cyclical heat treatments

Abstract Cu-based superelastic shape memory alloys are promising for low-stress elastocaloric cooling. We have synthesized bulk alloys of 68Cu–16Al–16Zn under different conditions in order to promote its grain growth and enhance its elastocaloric properties. High-temperature x-ray diffraction of untreated 68Cu–16Al–16Zn alloy showed that the phase boundary between the α + β mixed phases and the high temperature phase ( β phase) was between 973 K and 1023 K. Based on this result, the 68Cu–16Al–16Zn alloy was heated and cooled in a furnace repeatedly between 773 K and 1173 K. The maximum grain size after heat treatment of the ingot rolled to 67% reached 11.1 mm. The latent heat of the martensitic transformation after grain growth was 6.3 J g −1 , which is higher than the previously reported value for the compound. The stress–strain curve of 68Cu–16Al–16Zn rolled to 67% rolling with cyclical heat treatments showed a maximum stress of 106 MPa at 4.5% strain, with adiabatic temperature change of 5.9 K in heating during stress loading and 5.6 K in cooling in stress removal. Furthermore, no fatigue in the stress–strain behavior was observed up to at least 60 000 mechanical cycles at 2% strain.

36 MATERIALS SCIENCE↗
DOE OSTI2017

Resilient and Robust High Performance Computing Platforms for Scientific Computing Integrity

As technology advances, computer systems are subject to increasingly sophisticated cyber-attacks that compromise both their security and integrity. High performance computing platforms used in commercial and scientific applications involving sensitive, or even classified data, are frequently targeted by powerful adversaries. This situation is made worse by a lack of fundamental security solutions that both perform efficiently and are effective at preventing threats. Current security solutions fail to address the threat landscape and ensure the integrity of sensitive data. As challenges rise, both private and public sectors will require robust technologies to protect its computing infrastructure. The research outcomes from this project try to address all these challenges. For example, we present LAZARUS, a novel technique to harden kernel Address Space Layout Randomization (KASLR) against paging-based side-channel attacks. In particular, our scheme allows for fine-grained protection of the virtual memory mappings that implement the randomization. We demonstrate the effectiveness of our approach by hardening a recent Linux kernel with LAZARUS, mitigating all of the previously presented side-channel attacks on KASLR. Our extensive evaluation shows that LAZARUS incurs only 0.943% overhead for standard benchmarks, and is therefore highly practical. We also introduced HA2lloc, a hardware-assisted allocator that is capable of leveraging an extended memory management unit to detect memory errors in the heap. We also perform testing using HA2lloc in a simulation environment and find that the approach is capable of preventing common memory vulnerabilities.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2019

“Magnetic Memory” in Plasticity of an Aluminum Alloy with Iron Inclusions

This article is devoted to the effect of a magnetic field of 0.7 T on the state of an aluminum alloy whose creep increases after exposing to a magnetic field, exhibiting a “magnetic memory.” The alloy contains FeAl microinclusions identified by electron microscopy, magnetometry, and Mössbauer spectroscopy. Possible mechanisms of changing the states of ferromagnetic microinclusions and aluminum matrix around them under a magnetic field are discussed.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗
NASA NTRSDate not supplied

Early And Late Neurobehavioral Effects Of Male And Female Mice Exposed To Five-Ion GCRSim

With upcoming missions to the Moon and beyond to Mars, it is increasingly imperative to elucidate the detrimental effects of space travel beyond the lower Earth orbit, particularly to galactic cosmic radiation. Additionally, with the first female astronaut to soon travel to the Moon there is a strong need to understand the biological sex differences to adaptation to the deep space environment. While the effects of spaceflight on the nervous system are not fully known, studies in animal models have shown that exposure to ionizing radiation can cause neuronal damage and lead to downstream cognitive and behavioral deficits. Here, we investigated the neurobehavioral responses to space environment-like radiation exposure. Male and female 23–24-week-old mice (age-matched to average astronaut age) were exposed to 5, 15 and 50 cGy via Five-Ion Galactic Cosmic Ray Simulation at the NASA National Space Radiation Laboratory at Brookhaven National Laboratory. Both early (72hrs post exposure) and late (1-4 months post exposure) cognitive and behavioral deficits were investigated. Early analysis was performed by observing in-cage behavior including frequency/duration of digging, rearing, and grooming and nestlet building. Additionally, late effects were analyzed at NASA Ames via in-cage behavior as well as Catwalk (gait), Zero Maze (anxiety), Adhesive Removal (sensory motor), Novel Object Recognition and Barnes Maze (working memory). There were pronounced sex differences in both early and late time points. Females scored better than males in nestlet building, gait and working memory while males scored better at the sensory motor (adhesive removal) test. Dose effects were primarily observed at 50 cGy for both sexes, particularly with the Barnes Maze (working memory), where both males and females exposed to 50 cGy showed little to no improvement in test performance, though females again performed better than males. Currently, we are investigating the correlations between the observed behavior and cytokine levels in the plasma and brain. Future studies will continue to investigate cognitive consequences of galactic cosmic radiation in combination with other space-like environment stressors, including antigravity and social versus single housing.

S Puukila↗
DOE OSTI2023

Tuning In-Plane Magnetic Anisotropy and Interfacial Exchange Coupling in Epitaxial La 2/3 Sr 1/3 CoO 3 /La 2/3 Sr 1/3 MnO 3 Heterostructures

Controlling the in-plane magnetocrystalline anisotropy and interfacial exchange coupling between ferromagnetic (FM) layers plays a key role in next-generation spintronic and magnetic memory devices. In this work, we explored the effect of tuning the magnetocrystalline anisotropy of La 2/3 Sr 1/3 CoO 3 (LSCO) and La 2/3 Sr 1/3 MnO 3 (LSMO) layers and the corresponding effect on interfacial exchange coupling by adjusting the thickness of the LSCO layer (t LSCO ). The epitaxial LSCO/LSMO bilayers were grown on (110) o -oriented NdGaO 3 (NGO) substrates with a fixed LSMO (top layer) thickness of 6 nm and LSCO (bottom layer) thicknesses varying from 1 to 10 nm. Despite the small difference (~0.2%) in lattice mismatch between the two in-plane directions, [001] o and [11̅0] o , a pronounced in-plane magnetic anisotropy was observed. Soft X-ray magnetic circular dichroism hysteresis loops revealed that for t LSCO ≤ 4 nm, the easy axes for both LSCO and LSMO layers were along the [001] o direction, and the LSCO layer was characterized by magnetically active Co 2+ ions that strongly coupled to the LSMO layer. No exchange bias effect was observed in the hysteresis loops. In contrast, along the [11̅0] o direction, the LSCO and LSMO layers displayed a small difference in their coercivity values, and a small exchange bias shift was observed. As t LSCO increased above 4 nm, the easy axis for the LSCO layer remained along the [100] o direction, but it gradually rotated to the [11̅0] o direction for the LSMO layer, resulting in a large negative exchange bias shift. Therefore, we provide a way to control the magnetocrystalline anisotropy and exchange bias by tuning the interfacial exchange coupling between the two FM layers.

36 MATERIALS SCIENCE↗
DOE OSTI2015

Exact calculation of the time convolutionless master equation generator: Application to the nonequilibrium resonant level model

The generalized quantum master equation provides a powerful tool to describe the dynamics in quantum impurity models driven away from equilibrium. Two complementary approaches, one based on Nakajima–Zwanzig–Mori time-convolution (TC) and the other on the Tokuyama–Mori time-convolutionless (TCL) formulations provide a starting point to describe the time-evolution of the reduced density matrix. A key in both approaches is to obtain the so called “memory kernel” or “generator,” going beyond second or fourth order perturbation techniques. While numerically converged techniques are available for the TC memory kernel, the canonical approach to obtain the TCL generator is based on inverting a super-operator in the full Hilbert space, which is difficult to perform and thus, nearly all applications of the TCL approach rely on a perturbative scheme of some sort. Here, the TCL generator is expressed using a reduced system propagator which can be obtained from system observables alone and requires the calculation of super-operators and their inverse in the reduced Hilbert space rather than the full one. This makes the formulation amenable to quantum impurity solvers or to diagrammatic techniques, such as the nonequilibrium Green’s function. We implement the TCL approach for the resonant level model driven away from equilibrium and compare the time scales for the decay of the generator with that of the memory kernel in the TC approach. Furthermore, the effects of temperature, source-drain bias, and gate potential on the TCL/TC generators are discussed.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2019

Nanoscale magnetic phase competition throughout the Ni 50– x Co x Mn 40 Sn 10 phase diagram: Insights from small-angle neutron scattering

The Ni 2 MnSn-derived Ni 50– x Co x Mn 25+ y Sn 25– y alloys are premier examples of a class of off-stoichiometric Heusler alloys recently discovered to exhibit attractive magnetic properties in tandem with extraordinarily reversible martensitic phase transformations. Multiferroicity, magnetic phase competition and separation, field-induced martensitic transformations, magnetic shape memory behavior, and sizable magneto-, elasto-, and barocaloric effects result, generating substantial interest and application potential. In this work we expand on a prior small-angle neutron scattering (SANS) study at a single composition (Ni 44 Co 6 Mn 40 Sn 10 ) by exploring all three main regions of the recently established Ni 50–x Co x Mn 40 Sn 10 phase diagram, i.e., at the representative y = 15 composition. Wide temperature and scattering wave-vector range (20–500K, 0.004–0.2Å –1 ) SANS data on x = 2, 6, and 14 polycrystals provide a detailed picture of the evolution in magnetic order and inhomogeneity. Consistent with recent studies with a variety of techniques, phase separation into short-range coexisting ferromagnetic and antiferromagnetic regions is deduced below the martensitic transformation at x = 2 and 6, with average ferromagnetic cluster spacing of ~13 nm. Remarkably, at x = 14, where the martensitic transformation is suppressed and ferromagnetic austenite is stabilized to low temperatures, nanoscopic magnetic inhomogeneity nevertheless persists. Further, distinct ferromagnetic clusters (~36-nm average spacing) in a ferromagnetic matrix are observed at intermediate temperatures, homogenizing into a uniform long-range ordered ferromagnet only at low temperatures. This unusual ferromagnet cluster/ferromagnet matrix inhomogeneity, as well as x -dependent subtleties of the superparamagnetic freezing of ferromagnetic clusters, are discussed in light of 55 Mn nuclear magnetic resonance data, and the recent observation of annealing-induced core/shell nanoprecipitates. The origins of nanoscalemagnetic inhomogeneity are discussed in terms of statistical variations in local composition and structure, tendency to chemical phase separation, and other forms of disorder.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2024

Bridging the Gap Between LLMs and LNS with Dynamic Data Format and Architecture Codesign

Deep neural networks (DNNs) have achieved tremendous success in the past few years. However, their training and inference demand exceptional computational and memory resources. Quantization has been shown as an effective approach to mitigate the cost, with the mainstream data types reduced from FP32 to FP16/BF16 and recently FP8 in the latest NVIDIA H100 GPUs. With increasingly aggressive quantization, however, the conventional floating-point formats suffer from limited precision in representing numbers around zero. Recently, NVIDIA demonstrated the potential of using a Logarithmic Number System (LNS) for the next generation of tensor cores. While LNS mitigates the hurdles in representing small numbers, in this work we observed a mismatch between LNS and the emerging Large Language Models (LLM), where LLM exhibits significant outliers when directly adopting the LNS format. In this paper, we present a data-format/architecture codesign to bright this gap. On the format side, we propose a dynamic LNS format to flexibly represent outliers at a higher precision, by exploiting asymmetry in the LNS representation and identifying outliers through a per-vector basis. On the architecture side, for demonstration, we realize the dynamic LNS format in a systolic array, which can handle the irregularity of the outliers at runtime. We implement our approach on an Alveo U280 FPGA as a prototype. Experimental results show that our design can effectively handle the outliers and resolve the mismatch between LNS and LLM, contributing to an accuracy improvement of 15.4% and 16% over the floating-point and the original LNS baselines, using four state-of-the-art LLM models. Our observation and design lay a solid foundation for the large-scale adoption of the LNS format in the next-generation deep learning hardware.

Haghi, Pouya↗
DOE OSTI2016

Confronting Weather and Climate Models with Observational Data from Soil Moisture Networks over the United States

Four land surface models in uncoupled and coupled configurations are compared to observations of daily soil moisture from 19 networks in the conterminous United States to determine the viability of such comparisons and explore the characteristics of model and observational data. First, observations are analyzed for error characteristics and representation of spatial and temporal variability. Some networks have multiple stations within an area comparable to model grid boxes; for those it is found that aggregation of stations before calculation of statistics has little effect on estimates of variance, but soil moisture memory is sensitive to aggregation. Statistics for some networks stand out as unlike those of their neighbors, likely because of differences in instrumentation, calibration, and maintenance. Buried sensors appear to have less random error than near-field remote sensing techniques, and heat-dissipation sensors show less temporal variability than other types. Model soil moistures are evaluated using three metrics: standard deviation in time, temporal correlation (memory), and spatial correlation (length scale). Models do relatively well in capturing large-scale variability of metrics across climate regimes, but they poorly reproduce observed patterns at scales of hundreds of kilometers and smaller. Uncoupled land models do no better than coupled model configurations, nor do reanalyses outperform free-running models. Spatial decorrelation scales are found to be difficult to diagnose. Using data for model validation, calibration, or data assimilation from multiple soil moisture networks with different types of sensors and measurement techniques requires great caution. Finally, data from models and observations should be put on the same spatial and temporal scales before comparison.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2025

Space-Time Finite Element Tensor Network Approach for the Time-Dependent Convection–Diffusion–Reaction Equation with Variable Coefficients

In this paper, we present a new space-time Galerkin-like method, where we treat the discretization of spatial and temporal domains simultaneously. This method utilizes a mixed formulation of the tensor-train (TT) and quantized tensor-train (QTT) (please see Section Tensor-Train Decomposition), designed for the finite element discretization (Q1-FEM) of the time-dependent convection–diffusion–reaction (CDR) equation. We reformulate the assembly process of the finite element discretized CDR to enhance its compatibility with tensor operations and introduce a low-rank tensor structure for the finite element operators. Recognizing the banded structure inherent in the finite element framework’s discrete operators, we further exploit the QTT format of the CDR to achieve greater speed and compression. Additionally, we present a comprehensive approach for integrating variable coefficients of CDR into the global discrete operators within the TT/QTT framework. The effectiveness of the proposed method, in terms of memory efficiency and computational complexity, is demonstrated through a series of numerical experiments, including a semi-linear example.

convection–diffusion–reaction equation↗
NASA NTRS1973

On the error probability of general tree and trellis codes with applications to sequential decoding

An upper bound on the average error probability for maximum-likelihood decoding of the ensemble of random binary tree codes is derived and shown to be independent of the length of the tree. An upper bound on the average error probability for maximum-likelihood decoding of the ensemble of random L-branch binary trellis codes of rate R = 1/n is derived which separates the effects of the tail length T and the memory length M of the code. It is shown that the bound is independent of the length L of the information sequence. This implication is investigated by computer simulations of sequential decoding utilizing the stack algorithm. These simulations confirm the implication and further suggest an empirical formula for the true undetected decoding error probability with sequential decoding.

Johannesson, R.↗