Search NASA⌕ Search

SEARCH · Search NASA

Results for “Memory device”

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 1,513 records · Page 84

Experiences with SYCL on AMD GPUs with Kokkos

With the recent diversification of the hardware landscape in the high-performance computing (HPC) community, performance-portability solutions are becoming more and more important. One of the most popular choices is Kokkos, which recently became a Linux Foundation project. Most of its development is supported by the US Department of Energy and the French Alternative Energies and Atomic Energy Commission. Kokkos is implemented as a C++ library with multiple backends to support CPUs as well as various GPU architectures. These backends include OpenMP, CUDA, HIP, and also SCYL. This approach enables users to leverage the preferred vendor toolchain for the respective platform (e.g. CUDA, ROCm, OneAPI). The SYCL backend is used to target Intel GPUs, in particular to support the Aurora exascale supercomputer. However, SYCL itself also offers a large degree of portability, and in fact Kokkos’ CI for SYCL has been running on NVIDIA hardware due to a lack of access to Intel GPUs. In this report, we describe our experience with using Kokkos SYCL backend on AMD GPUs targeting the Frontier supercomputer at Oak Ridge National Laboratory. The two major SYCL implementations are DPC++ and AdaptiveCpp. While the Kokkos SYCL backend has been implemented using the former, the latter was the first implementation to target AMD GPUs. We will discuss the experience with both of these SYCL implementations in terms of functionality and performance. Using Kokkos to evaluate SYCL toolchains has a number of benefits. Kokkos’ use of SYCL is fairly complex, exercising features such as graphs, relocatable device functions, atomics – including for non-arithmetic types, as well as pinned and page migratable memory allocations. Kokkos also needs to implement capabilities such as Kokkos’ hierarchical parallelism that are not a straight-forward mapping to SYCL capabilities. Furthermore, a large number of libraries and applications that represent diverse use cases are implemented in Kokkos, providing readily available test cases for a toolchain evaluation. Preliminary results show that support for AMD GPUs in DPC++ is much less mature than for NVIDIA GPUs or Intel GPUs. While the situation has improved significantly over the last year, we still encounter many runtime failures, dispatching problems, and code generation issues. With AdaptiveCpp the challenges arise even earlier in the evaluation process. Since Kokkos’ SYCL implementation is largely focused on supporting Intel GPUs, we opted to leverage SYCL extensions which are available in DPC++ but not in AdaptiveCpp. Furthermore, AdaptiveCpp appears to be less conformant with the SYCL2020 standard which Kokkos relies on. In some cases, we are able to work around the lack of feature support, in other cases we have to disable certain Kokkos capabilities to evaluate the toolchain. Our evaluation will leverage Kokkos’ unit tests to establish basic functionality and feature completeness. We then use simple benchmarks for components of a CG implementation as a measure of usability and performance of the SYCL toolchains.

97 MATHEMATICS AND COMPUTING↗

Spacecraft optical disk recorder memory buffer control

The goal of this project is to develop an Application Specific Integrated Circuit (ASIC) for use in the control electronics of the Spacecraft Optical Disk Recorder (SODR). Specifically, this project is to design an extendable memory buffer controller ASIC for rate matching between a system Input/Output port and the SODR's device interface. The aforementioned goal can be partitioned into the following sub-goals: (1) completion of ASIC design and simulation (on-going via ASEE fellowship); (2) ASIC Fabrication (at ASIC manufacturer); and (3) ASIC Testing (NASA/LaRC, Christopher Newport University).

Hodson, Robert F.↗

Controlled fabrication of Si nanocrystal delta-layers in thin SiO{sub 2} layers by plasma immersion ion implantation for nonvolatile memories

Plasma Immersion Ion Implantation (PIII) is a promising alternative to beam line implantation to produce a single layer of nanocrystals (NCs) in the gate insulator of metal-oxide semiconductor devices. We report herein the fabrication of two-dimensional Si-NCs arrays in thin SiO{sub 2} films using PIII and rapid thermal annealing. The effect of plasma and implantation conditions on the structural properties of the NC layers is examined by transmission electron microscopy. A fine tuning of the NCs characteristics is possible by optimizing the oxide thickness, implantation energy, and dose. Electrical characterization revealed that the PIII-produced-Si NC structures are appealing for nonvolatile memories.

36 MATERIALS SCIENCE↗

Memory forensic analysis of a programmable logic controller in industrial control systems

In industrial control systems (ICS), programmable logic controllers (PLCs) are used to automate physical processes such as nuclear plants and power grid stations, and are often subject to cyber attacks. As in conventional IT domain, the memory analysis of the PLCs can help answer important forensic questions about the attack, such as the presence of malicious firmware, injection of modified control logic (the program running on the PLC), and manipulation of I/O devices (e.g., sensors and actuators). Unlike conventional IT domain, PLCs have heterogeneous hardware architecture, proprietary firmware and control software, making it challenging to employ a unified framework for their memory forensics. For merely extracting artifacts of forensic importance, reverse-engineering the firmware is a tedious task, and the effort needs to be repeated for every PLC model. As a community, a step-wise approach to tackle this challenge is to analyze the memory of specific PLCs, and subsequently find a generic framework applicable to all PLCs. Our work is a step forward in this direction. By following a methodology that focuses on the functional layer of PLCs instead of reverse engineering the firmware, we analyze the digital forensic artifacts available in a common PLC, Allen-Bradley ControlLogix 1756-L61. Before diving into the memory dump, we analyze the PLC control software to create a list of important artifacts that are sure to exist in the PLC memory dump. The approach employs a setup where PLC control software RSLogix-5000 is connected to the PLC, and the memory dump can be obtained as and when needed. We create test cases that sequentially highlight each category of artifacts, followed by an examination of the resultant impact on memory. After attaining the listed artifacts, we employ conventional string and known data searches to extract interesting information present in this PLC's memory. The memory analysis profile, presented as a Python library and shared with the community, can help a forensic investigator to readily extract forensic artifacts from the same model's controller. The adopted approach may help researchers in creating memory profile of other PLCs, and ultimately formulating a generic PLC memory analysis framework.

Rais, Muhammad Haris↗

Magnetic tunnel junctions with voltage tunable interlayer coupling for memory and sensor applications

Various examples are provided for magnetic tunnel junctions and applications thereof. In one example, a magnetic tunnel junction (MTJ) device includes a first ferromagnetic (FM) layer; a gadolinium oxide (GdOX) tunnel barrier disposed on the first ferromagnetic layer; and a second FM layer disposed on the GdOX tunnel barrier. In another example, a perpendicular MTJ (pMTJ) device includes a first layer including a magnetic material; a tunnel barrier disposed on the first layer to form the pMTJ; and a second layer including the magnetic material, the second layer disposed on the tunnel barrier.

Wang, Weigang↗

AYUSH: A Technique for Extending Lifetime of SRAM-NVM Hybrid Caches

Recently, researchers have explored way-based hybrid SRAM-NVM (non-volatile memory) last level caches (LLCs) to bring the best of SRAM and NVM together. However, the limited write endurance of NVMs restricts the lifetime of these hybrid caches. We present AYUSH, a technique to enhance the lifetime of hybrid caches, which works by using data-migration to preferentially use SRAM for storing frequently-reused data. Microarchitectural simulations confirm that AYUSH achieves larger improvement in lifetime than a previous technique and also maintains performance and energy efficiency. For single, dual and quad-core workloads, the average increase in cache lifetime with AYUSH is 6.90X, 24.06X and 47.62X, respectively.

device lifetime↗

Recent Trends and Innovation in Additive Manufacturing of Soft Functional Materials

The growing demand for wearable devices, soft robotics, and tissue engineering in recent years has led to an increased effort in the field of soft materials. With the advent of personalized devices, the one-shape-fits-all manufacturing methods may soon no longer be the standard for the rapidly increasing market of soft devices. Recent findings have pushed technology and materials in the area of additive manufacturing (AM) as an alternative fabrication method for soft functional devices, taking geometrical designs and functionality to greater heights. For this reason, this review aims to highlights recent development and advances in AM processable soft materials with self-healing, shape memory, electronic, chromic or any combination of these functional properties. Furthermore, the influence of AM on the mechanical and physical properties on the functionality of these materials is expanded upon. Additionally, advances in soft devices in the fields of soft robotics, biomaterials, sensors, energy harvesters, and optoelectronics are discussed. Lastly, current challenges in AM for soft functional materials and future trends are discussed.

36 MATERIALS SCIENCE↗

Field-free spin-orbit torque switching assisted by in-plane unconventional spin torque in ultrathin [Pt/Co] N

Electrical manipulation of magnetization without an external magnetic field is critical for the development of advanced non-volatile magnetic-memory technology that can achieve high memory density and low energy consumption. Several recent studies have revealed efficient out-of-plane spin-orbit torques (SOTs) in a variety of materials for field-free type-z SOT switching. Here, we report on the corresponding type-x configuration, showing significant in-plane unconventional spin polarizations from sputtered ultrathin [Pt/Co] N , which are either highly textured on single crystalline MgO substrates or randomly textured on SiO 2 coated Si substrates. The unconventional spin currents generated in the low-dimensional Co films result from the strong orbital magnetic moment, which has been observed by X-ray magnetic circular dichroism (XMCD) measurement. The x-polarized spin torque efficiency reaches up to −0.083 and favors complete field-free switching of CoFeB magnetized along the in-plane charge current direction. Micromagnetic simulations additionally demonstrate its lower switching current than type-y switching, especially in narrow current pulses. Our work provides additional pathways for electrical manipulation of spintronic devices in the pursuit of high-speed, high-density, and low-energy non-volatile memory.

42 ENGINEERING↗

About the deformation of ferroelectric hystereses

Studying ferroelectric hafnium oxide with focus on memory applications for the past years, discussions frequently involved the shape of measured polarization hystereses, its relation to the device performance, and how to optimize it. A perfect model-like hysteresis is of nearly rectangular shape and all deviations from this situation have to have a certain physical origin. Different phenomena and their impact on the shape of the polarization hystereses were reported in literature: Aging, imprint, fatigue, or dielectric interface layers to name a few examples. A collection of these phenomena is not easily found up to now. Thus, filling or at least reducing this gap is one of the goals of this work. Moreover, observing a pinched, slanted, or displaced hysteresis, it is quite tempting to try the reverse approach: a derivation of potential structural origins for this curve shape. First, the basics of the dynamic hysteresis measurement and the ferroelectric memories are briefly reviewed. The figures of interest are derived to ensure a proper assessment of imperfections in the hysteresis shape and their influence. It is discussed how a closer look on the polarization loop helps to draw conclusions on what might have caused such a shape or at least how to rule out some phenomena if the expected indications are not reflected in the present curve. Of course, further structural or electrical studies, as they are exemplarily pointed out, are indispensable to find the root cause(s) for the deviations from the ideal hysteresis. But sophisticated methods are not always accessible straightaway and, moreover, a pointer on where to start is always helpful. Especially, the transient currents recorded during a dynamic hysteresis measurement are stressed as a valuable instrument for this purpose. Despite their known potential, these currents are seldom shown in literature.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗

Micron MT29F128G08AJAAA 128GB Asynchronous Flash Memory Total Ionizing Dose Characterization Test Report

The purpose of this test was to characterize the Micron MT29F128G08AJAAAs parameter degradation for total dose response and to evaluate and compare lot date codes for sensitivity. In the test, the device was exposed to both low dose and high dose rate (HDR) irradiations using gamma radiation. Device parameters such as leakage currents, quantity of upset bits and overall chip and die health were investigated to determine which lot is more robust.

Flash Memory↗

Tunable Interfacial to Filamentary Resistive Switching Mechanism in Room-Temperature-Grown Amorphous YBa 2 Cu 3 O x with Excess Cu Addition

Resistive switching technologies have the potential not only to create large efficiency gains in computer memory but also to revolutionize emerging fields such as neuromorphic computing. In this paper, we report on novel resistive switching behavior in devices made from room-temperature-grown Cu-rich amorphous YBa 2 Cu 3 O x (YBCO) films, a material otherwise well-known as a high-temperature superconductor. In Nb:STO substrate/amorphous YBCO film (≈200 nm)/metallic Cu (15 nm)/metallic Pt (15 nm) devices, we demonstrate that the resistive switching can be tuned between mechanisms involving extended areas of the YBCO/electrode interface and a single-point filamentary mechanism simply by changing the Cu content of the deposition target and hence in the films. Changing the Cu content can also be used to optimize the properties of the devices further, with devices with an added 15 mol % of Cu in YBCO initially providing an on/off ratio >100, switching endurance potential >6500 cycles, and state retention >2 × 10 4 s, all at low switching fields of 0.3 MV/cm. The amalgam of promising resistive switching properties, fast growth (150 nm/min) at room temperature, and tuneability of the switching mechanism indicates the strong potential of this proof-of-concept amorphous system for future memory applications.

Cu↗

A Comprehensive Study on Energy Efficiency and Performance of Flash-based SSD

Use of flash memory as a storage medium is becoming popular in diverse computing environments. However, because of differences in interface, flash memory requires a hard-disk-emulation layer, called FTL (flash translation layer). Although the FTL enables flash memory storages to replace conventional hard disks, it induces significant computational and space overhead. Despite the low power consumption of flash memory, this overhead leads to significant power consumption in an overall storage system. In this paper, we analyze the characteristics of flash-based storage devices from the viewpoint of power consumption and energy efficiency by using various methodologies. First, we utilize simulation to investigate the interior operation of flash-based storage of flash-based storages. Subsequently, we measure the performance and energy efficiency of commodity flash-based SSDs by using microbenchmarks to identify the block-device level characteristics and macrobenchmarks to reveal their filesystem level characteristics.

99 GENERAL AND MISCELLANEOUS↗

Automotive Li-Ion Batteries: Current Status and Future Perspectives

Lithium-ion batteries (LIBs) are currently the most suitable energy storage device for powering electric vehicles (EVs) owing to their attractive properties including high energy efficiency, lack of memory effect, long cycle life, high energy density and high power density. These advantages allow them to be smaller and lighter than other conventional rechargeable batteries such as lead–acid batteries, nickel–cadmium batteries (Ni–Cd) and nickel–metal hydride batteries (Ni–MH). Modern EVs, however, still suffer from performance barriers (range, charging rate, lifetime, etc.) and technological barriers (high cost, safety, reliability, etc.), limiting their widespread adoption. Given these facts, this review sets the extensive market penetration of LIB-powered EVs as an ultimate objective and then discusses recent advances and challenges of electric automobiles, mainly focusing on critical element resources, present and future EV markets, and the cost and performance of LIBs. Finally, novel battery chemistries and technologies including high-energy electrode materials and all-solid-state batteries are also evaluated for their potential capabilities in next-generation long-range EVs.

25 ENERGY STORAGE↗

Weighted Traffic Equilibrium Problem in Non Pivot Hilbert Spaces with Long Term Memory

In the paper we consider a weighted traffic equilibrium problem in a non-pivot Hilbert space and prove the equivalence between a weighted Wardrop condition and a variational inequality with long term memory. As an application we show, using recent results of the Senseable Laboratory at MIT, how wireless devices can be used to optimize the traffic equilibrium problem.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Flexible spin-orbit torque devices

We report on state-of-the-art spintronic devices synthesized and fabricated directly on a flexible organic substrate. Large perpendicular magnetic anisotropy was achieved in ultrathin ferromagnetic heterostructures of Pt/Co/MgO sputtered on a non-rigid plastic substrate at room temperature. Subsequently, a full magnetic reversal of the Co was observed by exploiting the spin orbit coupling in Pt that leads to a spin accumulation at the Pt/Co interface when an in-plane current is applied. Quasi-static measurements show the potential for operating these devices at nano-second speeds. Importantly, the behavior of the devices remained unchanged under varying bending conditions (up to a bending radius of ≈ ±20–30 mm). Furthermore, the devices showed robust operation even after application of 10{sup 6} successive pulses, which is likely sufficient for many flexible applications. Thus, this work demonstrates the potential for integrating high performance spintronic devices on flexible substrates, which could lead to many applications ranging from flexible non-volatile magnetic memory to local magnetic resonance imaging.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Towards Compact Neural Networks via End-to-End Training: A Bayesian Tensor Approach with Automatic Rank Determination

Post-training model compression can reduce the inference costs of deep neural networks, but uncompressed training still consumes enormous hardware resources and energy. To enable low-energy training on edge devices, it is highly desirable to directly train a compact neural network from scratch with a low memory cost. Low-rank tensor decomposition is an effective approach to reduce the memory and computing costs of large neural networks. However, directly training low-rank tensorized neural networks is a very challenging task because it is hard to determine a proper tensor rank a priori, and the tensor rank controls both model complexity and accuracy. Here, this paper presents a novel end-to-end framework for low-rank tensorized training. We first develop a Bayesian model that supports various low-rank tensor formats (e.g., CANDECOMP/PARAFAC, Tucker, tensor-train, and tensor-train matrix) and reduces neural network parameters with automatic rank determination during training. Then we develop a customized Bayesian solver to train large-scale tensorized neural networks. Our training methods shows orders-of-magnitude parameter reduction and little accuracy loss (or even better accuracy) in the experiments. On a very large deep learning recommendation system with over 4.2 ×10 9 model parameters, our method can reduce the parameter number to 1.6 ×10 5 automatically in the training process (i.e., by 2.6 ×10 4 times) while achieving almost the same accuracy. Code is available at https://github.com/colehawkins/bayesian-tensor-rank-determination.

compact neural networks↗