Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Stacked 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 217 records · Page 12

Multilevel and long retentive resistive switching in low temperature nanostructured Cu/SiO{sub x}-W-SiO{sub x}/Pt

Amorphous SiO{sub x}-based memory films are fabricated at room temperature, and study on their resistive switching characteristics and improvement approaches is performed. Multilevel resistive states with large ratio 1: ∼10{sup 2}: 3 × 10{sup 5} and long retention exceeding 2 × 10{sup 6} s at ambient temperature and humidity are observed in Cu/SiO{sub x} (9 nm)-W (∼2 nm)-SiO{sub x} (9 nm)/Pt ultrathin stack. Nonvolatile switching is consistently realized in microscopy. Based on investigations of microscopic conduction and microstructure, tungsten incorporation with copper as relay bridges for conducting filaments is proposed to attribute to the performance improvement and the multilevel switching mechanism.

77 NANOSCIENCE AND NANOTECHNOLOGY↗

Supporting the future nuclear workforce with computer-based procedures

Here we see that computer-based tools have dramatically increased ease and efficiency of everyday tasks. Gone are the days of paging through a paper catalog, transcribing product numbers, and calculating totals. Today, a consumer can find a product online with a simple search engine, and then purchase it in a matter of a few clicks. Paper catalogs have their place, but it is hard to imagine life without on-line shopping sites. All tasks conducted in a nuclear power plant are guided by procedures, which helps ensure safe and reliable operation of the plants. One prominent goal of the nuclear industry is to minimize the risk of human errors. To achieve this goal one has to ensure tasks are correctly and consistently executed. This is partly achieved by training and by a structured approach to task execution, which is provided by procedures and work instructions. Procedures are used in the nuclear industry to direct workers' actions in a proper sequence. The governing idea is to minimize the reliance on memory and choices made in the field. However, the procedure document may not contain sufficient information to successfully complete the task. Therefore, the worker might have to carry additional documents such as turnover sheets, operation experience, drawings, and other procedures to the work site. The nuclear industry is operated with paper procedures like paper catalogs of the past. A field worker may carry a large stack of documents needed to complete a task to the field. Even though the paper process has helped keep the industry safe for decades, there are limitations to using paper. Paper procedures are static (i.e., the content does not change after the document is printed), difficult to search, and rely heavily on the field worker’s situational awareness and ability to consistently meet the high expectation of human performance excellence. With computer-based procedures (CBPs) that stack of papers may be reduced to the size of a small tablet or even a smart phone. Instead of manually matching equipment identification numbers listed in the procedure with the number on the physical equipment the field worker can simply scan a barcode to ensure the correct valve is opened while simultaneously creating a record. Instead of navigating through a maze of cross-references, CBPs enable intelligent work path navigation which accounts for past decisions and observation, thereby enabling more efficient and safe task completion.

47 OTHER INSTRUMENTATION↗

Performance Study of the First 2D Prototype of Vertically Integrated Pattern Recognition Associative Memory (VIPRAM)

Extremely fast pattern recognition capabilities are necessary to find and fit billions of tracks at the hardware trigger level produced every second anticipated at high luminosity LHC (HL-LHC) running conditions. Associative Memory (AM) based approaches for fast pattern recognition have been proposed as a potential solution to the tracking trigger. However, at the HL-LHC, there is much less time available and speed performance must be improved over previous systems while maintaining a comparable number of patterns. The Vertically Integrated Pattern Recognition Associative Memory (VIPRAM) Project aims to achieve the target pattern density and performance goal using 3DIC technology. The first step taken in the VIPRAM work was the development of a 2D prototype (protoVIPRAM00) in which the associative memory building blocks were designed to be compatible with the 3D integration. In this paper, we present the results from extensive performance studies of the protoVIPRAM00 chip in both realistic HL-LHC and extreme conditions. Results indicate that the chip operates at the design frequency of 100 MHz with perfect correctness in realistic conditions and conclude that the building blocks are ready for 3D stacking. We also present performance boundary characterization of the chip under extreme conditions.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

Probabilistic Nanomagnetic Memories for Uncertain and Robust Machine Learning

This project evaluated the use of emerging spintronic memory devices for robust and efficient variational inference schemes. Variational inference (VI) schemes, which constrain the distribution for each weight to be a Gaussian distribution with a mean and standard deviation, are a tractable method for calculating posterior distributions of weights in a Bayesian neural network such that this neural network can also be trained using the powerful backpropagation algorithm. Our project focuses on domain-wall magnetic tunnel junctions (DW-MTJs), a powerful multi-functional spintronic synapse design that can achieve low power switching while also opening the pathway towards repeatable, analog operation using fabricated notches. Our initial efforts to employ DW-MTJs as an all-in-one stochastic synapse with both a mean and standard deviation didn’t end up meeting the quality metrics for hardware-friendly VI. In the future, new device stacks and methods for expressive anisotropy modification may make this idea still possible. However, as a fall back that immediately satisfies our requirements, we invented and detailed how the combination of a DW-MTJ synapse encoding the mean and a probabilistic Bayes-MTJ device, programmed via a ferroelectric or ionically modifiable layer, can robustly and expressively implement VI. This design includes a physics-informed small circuit model, that was scaled up to perform and demonstrate rigorous uncertainty quantification applications, up to and including small convolutional networks on a grayscale image classification task, and larger (Residual) networks implementing multi-channel image classification. Lastly, as these results and ideas all depend upon the idea of an inference application where weights (spintronic memory states) remain non-volatile, the retention of these synapses for the notched case was further interrogated. These investigations revealed and emphasized the importance of both notch geometry and anisotropy modification in order to further enhance the endurance of written spintronic states. In the near future, these results will be mapped to effective predictions for room temperature and elevated operation DW-MTJ memory retention, and experimentally verified when devices become available.

77 NANOSCIENCE AND NANOTECHNOLOGY↗

Twist-assisted all-antiferromagnetic tunnel junction in the atomic limit

Abstract Antiferromagnetic spintronics 1,2 shows great potential for high-density and ultrafast information devices. Magnetic tunnel junctions (MTJs), a key spintronic memory component that are typically formed from ferromagnetic materials, have seen rapid developments very recently using antiferromagnetic materials 3,4 . Here we demonstrate a twisting strategy for constructing all-antiferromagnetic tunnel junctions down to the atomic limit. By twisting two bilayers of CrSBr, a 2D antiferromagnet (AFM), a more than 700% nonvolatile tunnelling magnetoresistance (TMR) ratio is shown at zero field (ZF) with the entire twisted stack acting as the tunnel barrier. This is determined by twisting two CrSBr monolayers for which the TMR is shown to be derived from accumulative coherent tunnelling across the individual CrSBr monolayers. The dependence of the TMR on the twist angle is calculated from the electron-parallel momentum-dependent decay across the twisted monolayers. This is in excellent agreement with our experiments that consider twist angles that vary from 0° to 90°. Moreover, we also find that the temperature dependence of the TMR is, surprisingly, much weaker for the twisted as compared with the untwisted junctions, making the twisted junctions even more attractive for applications. Our work shows that it is possible to push nonvolatile magnetic information storage to the atomically thin limit.

Science & Technology - Other Topics↗

On the Marriage of Asynchronous Many Task Runtimes and Big Data: A Glance

The rise of the accelerator-based architectures and reconfigurable computing have showcased the weakness of software stack toolchains that still maintain a static view of the hardware instead of relying on a symbiotic relationship between static (e.g., compilers) and dynamic tools (e.g., runtimes). In the past decades, this need has given rise to adaptive runtimes with increasingly finer computational tasks. These finer tasks help to take advantage of the hardware by switching out when a long latency operation is encountered (because of the deeper memory hierarchies and new memory technologies that might target streaming instead of random access), thus trading off idle time for unrelated work. Examples of these finer task runtimes are Asynchronous Many Task (AMT) runtimes, in which highly efficient computational graphs run on a variety of hardware. Due to its inherent latency tolerant characteristics, Latency-sensitive applications, such as Graph Analytics and Big Data can effectively use these runtimes. This paper aims to present an example of how the careful design of an AMT can exploit the hardware substrate when faced with high latency applications such as the ones given in the Big Data domain. Moreover, with its introspection and adaptive capabilities, we aim to show the power of these runtimes when facing the changing requirements of the application workloads. We use the Performance Open Community Runtime (P-OCR) as our vehicle to demonstrate the concepts presented here.

adaptive runtime, big data analysis↗

Effect of Annealing on Perpendicular Magnetic Anisotropy and Low Saturation Magnetization of MgO/Co{sub 2}FeAl/Mo Trilayer Films

In this study, we report our approach to manipulate the strength of perpendicular magnetic anisotropy (PMA) at oxide/ferromagnetic metal of Heusler alloy (HA) interface by tuning the parameter of annealing temperature. Robust PMA energy density (K{sub eff}) of ∼ 0.43 × 10{sup 6} erg/cm{sup 3} and small saturation magnetization (M{sub s}) as low as ∼ 248 emu/cm{sup 3} have been achieved for MgO/Co{sub 2}FeAl (CFA)/Mo stacks at 400 {sup ∘}C. We identify that PMA energy is thermally stable up to the maximum temperature of 400 {sup ∘}C. Effect of annealing temperature on surface morphology, grain distribution, and surface roughness of the stacks are investigated by atomic force microscopy (AFM). Further, X-ray photoelectron spectroscopy (XPS) study demonstrates that the interfacial oxidation decides the strong PMA property at MgO/Co{sub 2}FeAl interface during the annealing process. This MgO/CFA/Mo trilayer system can be a suitable candidate for the development of low-power consumption, huge-density data storage, high-thermal stability against the annealing temperature, and nonvolatile magnetic memories.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗

Userspace Squash Filesystem for Launching Linux Containers on HPC Systems [Thesis]

The demand for user defined software stacks (UDSS) has been increasing in the high-performance computing (HPC) community. Container technology has become popular due to the flexibility and isolation it provides to HPC users. Container images must be available to all nodes involved for use in HPC and can be distributed to compute nodes in a variety of ways. A common method for container image distribution is to simply copy the container image to memory on each compute node, which can be time consuming at scale, and uses valuable memory on each node. The kernel mounted squash filesystem (squashfs) has proven fast and efficient for this task but requires root-level access. This paper will show a user space mounted squashfs is an efficient and secure solution for container image distribution.

97 MATHEMATICS AND COMPUTING↗

Vertical 2T-nC FeRAM Demonstration: BEOL Read Transistor for 4F 2 Memory Strings and Two-Terminal Selector Design for Polarization Disturb Mitigation

In this work, we demonstrate a vertical 2T-nC FeRAM with a back-end-of-line (BEOL) read transistor (T R ) for 4F 2 string and propose a selector design to mitigate polarization disturb in passive capacitor crossbar arrays. Key contributions include: 1) successful integration and operation composed of a Si MOSFET write transistor (T W ), 3-layer cylindrical ferroelectric capacitors, and Si-doped In 2 O 3 BEOL T R , demonstrating the feasibility of 4F 2 2T−nC string; 2) introducing nonlinearity into the capacitor stack to suppress ferroelectric voltage drop under inhibition biases while maintaining sufficient write voltage, reducing disturbance; 3)modeling and experimental validation of inserting a metalsemiconductor (a-Si)-metal (MSM) selector into the capacitor in mitigating the disturb, particularly achieving 9x reduction of disturb after 10 6 cycles in the V W /2 scheme.

42 ENGINEERING↗

NiH2 Battery Reconditioning for LEO Applications

This paper summarizes reasons for and benefits of reconditioning nickel-hydrogen (NiH2) batteries used for Low Earth Orbit (LEO) applications. NiH2 battery cells do not have the classic discharge voltage problems more commonly associated with nickel-cadmium (NiCd) cells. This is due, in part, to use of hydrogen electrodes in place of cadmium electrodes. The nickel electrode, however, does have a similar discharge voltage signature for both cell designs. This can have an impact on LEO applications where peak loads at higher relative depths of discharge can impact operations. Periodic reconditioning provides information which can be used for analyzing long term performance trends to predict usable capacity to a specified voltage level. The reconditioning process described herein involves discharging NiH2 batteries at C/20 rates or less, to an average cell voltage of 1.0 volts or less. Recharge is performed at nominal C/5 rates to specified voltage/temperature (V/T) charge levels selected to restore required capacity with minimal overcharge. Reconditioning is a process of restoring reserve capacity lost on cycling, which is commonly called the memory effect in NiCd cells. This effect is characterized by decreases in the discharge voltage curve with operational life and cycling. The end effect of reconditioning NiH2 cells may be hidden in the versatility, of that design over the NiCd cell design and its associated negative electrode fading problem. The process of deep discharge at lower rates by way of reconditioning tends to redistribute electrolyte and water in the NiH2 cell electrode stack, while improving utilization and charge efficiency. NiH2 battery reconditioning effects on life are considered beneficial and may, in fact. extend life based on NiCd experience. In any case, usable capacity data obtained from reconditioning is required for performance evaluation and trend analysis. Characterization and life tests have provided the historical data base used to determine the need for reconditioning in most battery applications. The following sections briefly describe the background of NiH2 battery reconditioning and testing at Lockheed Martin Missiles & Space (LMMS) and other aerospace companies.

Armantrout, J. D.↗

Design and implementation of I/O performance prediction scheme on HPC systems through large-scale log analysis

Abstract Large-scale high performance computing (HPC) systems typically consist of many thousands of CPUs and storage units used by hundreds to thousands of users simultaneously. Applications from large numbers of users have diverse characteristics, such as varying computation, communication, memory, and I/O intensity. A good understanding of the performance characteristics of each user application is important for job scheduling and resource provisioning. Among these performance characteristics, I/O performance is becoming increasingly important as data sizes rapidly increase and large-scale applications, such as simulation and model training, are widely adopted. However, predicting I/O performance is difficult because I/O systems are shared among all users and involve many layers of software and hardware stack, including the application, network interconnect, operating system, file system, and storage devices. Furthermore, updates to these layers and changes in system management policy can significantly alter the I/O behavior of applications and the entire system. To improve the prediction of the I/O performance on HPC systems, we propose integrating information from several different system logs and developing a regression-based approach to predict the I/O performance. Our proposed scheme can dynamically select the most relevant features from the log entries using various feature selection algorithms and scoring functions, and can automatically select the regression algorithm with the best accuracy for the prediction task. The evaluation results show that our proposed scheme can predict the write performance with up to 90% prediction accuracy and the read performance with up to 99% prediction accuracy using the real logs from the Cori supercomputer system at NERSC.

97 MATHEMATICS AND COMPUTING↗

RoCE: Promising Technology for Ethernet as a High Performance Networking Fabric

Remote Direct Memory Access (RDMA) over Converged Ethernet (RoCE) has the potential to provide performance that rivals traditional high performance fabrics. If this potential proves out, significant impacts on system procurement decisions could follow. This work provides a series of small scale performance results which are used to compare and contrast the performance of RoCE-enabled Ethernet with TCP-based Ethernet and an HPC network. Additionally, a discussion of the maturity of RoCE firmware/software stacks and documentation is provided along with useful approaches for probing performance. A detailed description of two experimental setups known to have good RoCE performance is given, including step-by-step configuration and the exact hardware and software revisions employed. At small scales, RoCE is found to have significant performance advantages over "out-of-the-box" TCP protocols and is competitive with state-of-the-art high performance networks. Further examination of RoCE using a wider array of benchmarks and at greater scale is warranted.

97 MATHEMATICS AND COMPUTING↗

Probabilistic Neural Computing with Stochastic Devices

Abstract The brain has effectively proven a powerful inspiration for the development of computing architectures in which processing is tightly integrated with memory, communication is event‐driven, and analog computation can be performed at scale. These neuromorphic systems increasingly show an ability to improve the efficiency and speed of scientific computing and artificial intelligence applications. Herein, it is proposed that the brain's ubiquitous stochasticity represents an additional source of inspiration for expanding the reach of neuromorphic computing to probabilistic applications. To date, many efforts exploring probabilistic computing have focused primarily on one scale of the microelectronics stack, such as implementing probabilistic algorithms on deterministic hardware or developing probabilistic devices and circuits with the expectation that they will be leveraged by eventual probabilistic architectures. A co‐design vision is described by which large numbers of devices, such as magnetic tunnel junctions and tunnel diodes, can be operated in a stochastic regime and incorporated into a scalable neuromorphic architecture that can impact a number of probabilistic computing applications, such as Monte Carlo simulations and Bayesian neural networks. Finally, a framework is presented to categorize increasingly advanced hardware‐based probabilistic computing technologies.

Misra, Shashank↗

Spin-orbit torque magnetization switching of a three-terminal perpendicular magnetic tunnel junction

We report on the current-induced magnetization switching of a three-terminal perpendicular magnetic tunnel junction by spin-orbit torque and its read-out using the tunnelling magnetoresistance (TMR) effect. The device is composed of a perpendicular Ta/FeCoB/MgO/FeCoB stack on top of a Ta current line. The magnetization of the bottom FeCoB layer can be switched reproducibly by the injection of current pulses with density 5 × 10{sup 11} A/m{sup 2} in the Ta layer in the presence of an in-plane bias magnetic field, leading to the full-scale change of the TMR signal. Our work demonstrates the proof of concept of a perpendicular spin-orbit torque magnetic memory cell.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗

Big Data Meets HPC Log Analytics: Scalable Approach to Understanding Systems at Extreme Scale

Today's high-performance computing (HPC) systems are heavily instrumented, generating logs containing information about abnormal events, such as critical conditions, faults, errors and failures, system resource utilization, and about the resource usage of user applications. These logs, once fully analyzed and correlated, can produce detailed information about the system health, root causes of failures, and analyze an application's interactions with the system, providing valuable insights to domain scientists and system administrators. However, processing HPC logs requires a deep understanding of hardware and software components at multiple layers of the system stack. Moreover, most log data is unstructured and voluminous, making it more difficult for system users and administrators to manually inspect the data. With rapid increases in the scale and complexity of HPC systems, log data processing is becoming a big data challenge. This paper introduces a HPC log data analytics framework that is based on a distributed NoSQL database technology, which provides scalability and high availability, and the Apache Spark framework for rapid in-memory processing of the log data. The analytics framework enables the extraction of a range of information about the system so that system administrators and end users alike can obtain necessary insights for their specific needs. We describe our experience with using this framework to glean insights from the log data about system behavior from the Titan supercomputer at the Oak Ridge National Laboratory.

Park, Byung↗

The Darwin Cluster

Darwin is an ASC funded test bed cluster that also allows external users to use the idle resources within. Darwin is configured as a non-standard HPC production stack, and as such, has some interesting quirks to it. Darwin is a very heterogeneous cluster. Where average clusters are designed as homogeneous to make management and expectations simple, Darwin was designed as a test bed with many types of nodes available for running. With this heterogeneity comes its own run time experience. We provide nodes that contain x86 architectures of many flavors and also Power PC and ARM architectures. We have nodes with terabytes of memory and nodes with all kinds of GPUs.

97 MATHEMATICS AND COMPUTING↗

MPACT 4.4 Theory Manual

MPACT is a three-dimensional (3D) full-core neutron transport code capable of calculating subpin power distributions. Calculations are based on the Boltzmann transport equation for neutron fluxes for problems in which the detailed geometrical configuration of fuel components such as the pellet and cladding are explicitly retained. The cross-section data needed for the neutron transport calculation are obtained directly from a multigroup cross section library, which has traditionally been used by lattice physics codes to generate few-group homogenized cross sections for nodal core simulators. Hence, MPACT assumes neither a priori homogenization nor group condensation for the full core spatial solution. The 3D MPACT transport solution can be obtained using the method of characteristics (MOC), which employs discrete ray tracing within each fuel pin. However, for practical reactor applications, the direct application of MOC to 3D core configurations requires an excessive amount of memory and computing time due to the very large number of rays. For practical 3D full-core calculations, MPACT commonly uses an approximate “2D/1D” method that treats the radial (x and y) variables differently from the axial (z) variable. In particular, the radial dependence of the solution is calculated using transport theory, and the axial dependence is calculated using diffusion or P 3 theory. The 2D/1D method requires the core to be divided into a vertical stack of axial slices with a thickness of Δ z ≈ 5–10 cm. Each axial slice is divided radially into coarse spatial cells with boundaries that usually constitute the pin cell boundaries, for which Δ x = Δ y ≈ 1.5 cm. Then, each coarse radial cell (pin cell) is divided into 50–100 fine radial cells, which resolve the angular flux in the fuel, cladding, and moderator regions.

22 GENERAL STUDIES OF NUCLEAR REACTORS↗

BEOL compatible high tunnel magneto resistance perpendicular magnetic tunnel junctions using a sacrificial Mg layer as CoFeB free layer cap

Perpendicularly magnetized MgO-based tunnel junctions are envisaged for future generation spin-torque transfer magnetoresistive random access memory devices. Achieving a high tunnel magneto resistance and preserving it together with the perpendicular magnetic anisotropy during BEOL CMOS processing are key challenges to overcome. The industry standard technique to deposit the CoFeB/MgO/CoFeB tunnel junctions is physical vapor deposition. In this letter, we report on the use of an ultrathin Mg layer as free layer cap to protect the CoFeB free layer from sputtering induced damage during the Ta electrode deposition. When Ta is deposited directly on CoFeB, a fraction of the surface of the CoFeB is sputtered even when Ta is deposited with very low deposition rates. When depositing a thin Mg layer prior to Ta deposition, the sputtering of CoFeB is prevented. The ultra-thin Mg layer is sputtered completely after Ta deposition. Therefore, the Mg acts as a sacrificial layer that protects the CoFeB from sputter-induced damage during the Ta deposition. The Ta-capped CoFeB free layer using the sacrificial Mg interlayer has significantly better electrical and magnetic properties than the equivalent stack without protective layer. We demonstrate a tunnel magneto resistance increase up to 30% in bottom pinned magnetic tunnel junctions and tunnel magneto resistance values of 160% at resistance area product of 5 Ω.μm{sup 2}. Moreover, the free layer maintains perpendicular magnetic anisotropy after 400 °C annealing.

36 MATERIALS SCIENCE↗