Carbon nanotube network-silicon oxide non-volatile switches
Explore the source record for details and available documents.
SEARCH · Search NASA
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.
Explore the source record for details and available documents.
Recently, a Cambrian explosion of novel, non-volatile memory (NVM) devices known as memristive devices have inspired effort in building hardware neural networks that learn like the brain. Early experimental prototypes built simple perceptrons from nanosynapses, and recently, fully-connected multi-layer perceptron (MLP) learning systems have been realized. However, while backpropagating learning systems pair well with high-precision computer memories and achieve state-of-the-art performances, this typically comes with a massive energy budget. For future Internet of Things/peripheral use cases, system energy footprint will be a major constraint, and emerging NVM devices may fill the gap by sacrificing high bit precision for lower energy. In this work, we contrast the well known MLP approach with the Extreme Learning Machine (ELM) or NoProp approach, which uses a large layer of random weights to improve the separability of high-dimensional tasks, and is usually considered inferior in a software context. However, we find that when taking device non-linearity into account, NoProp manages to equal hardware MLP system in terms of accuracy. While also using a sign-based adaptation of the delta rule for energy-savings, we find that NoProp can learn effectively with four to six ’bits’ of device analog capacity, while MLP requires eight bit capacity with the same rule. This may allow the requirements for memristive devices to be relaxed in the context of online learning. By comparing the energy footprint of these systems for several candidate nanosynapses, as well as realistic peripherals, we confirm that memristive NoProp systems save energy compared to MLP systems. Lastly, we show that ELM/NoProp systems can achieve better generalization abilities than nanosynaptic MLP systems when paired with pre-processing layers (which do not require backpropagated error). Collectively, these advantages make such systems worthy of consideration in future accelerators or embedded hardware.
The volatile organic compounds (VOCs) in a wide spectrum of categories were identified as biomarkers in aquatic environments, playing an important role in marine and freshwater ecology and global atmospheric chemistry. VOCs released from biofuel have also attracted increasing attention. Although the importance has been recognized, the portable detection and analysis methods of VOC in aquatic systems have not yet been well developed and understood. Here, in this work, we innovatively proposed an intelligent electrochemical sensing approach to classify and quantify VOCs in solution. Utilizing the cyclic voltammetry (CV) method with an ionic liquid (IL)-based electrolyte, we analyzed 50 μL samples of various VOC analytes, including acetic acid (AC), acetone, dimethylformamide (DMF), dimethyl sulfoxide (DMSO), ethanol, formaldehyde, formic acid, methanol, methyl formate (MF), toluene, and a formaldehyde-methanol mixture, along with deionized water (DI water). The generated voltammograms were subsequently analyzed using our uniquely designed and optimized 1-D convolutional neural network (1D-CNN). This deep-learning algorithm achieved a 99.09% accuracy in VOC classification validated through fivefold cross-validation and demonstrated an impressive 94.4% test accuracy for methanol detection within a 10 μL error range. For quantification, the system accurately categorized methanol volumes ranging from 0 to 50 μL in 10 μL increments, achieving a 98.18% accuracy. A notable linear correlation (R2 = 95.56%) was found between max current density at the oxidation peak and methanol volume, with the limit of detection (LOD) at 9.3 μL. Such a sensing method exhibits potential for portability, high accuracy, and generalization in the classification and quantification, ultimately reshaping the realm of VOC analysis in solution.
Abstract Deep learning has become ubiquitous, touching daily lives across the globe. Today, traditional computer architectures are stressed to their limits in efficiently executing the growing complexity of data and models. Compute‐in‐memory (CIM) can potentially play an important role in developing efficient hardware solutions that reduce data movement from compute‐unit to memory, known as the von Neumann bottleneck. At its heart is a cross‐bar architecture with nodal non‐volatile‐memory elements that performs an analog multiply‐and‐accumulate operation, enabling the matrix‐vector‐multiplications repeatedly used in all neural network workloads. The memory materials can significantly influence final system‐level characteristics and chip performance, including speed, power, and classification accuracy. With an over‐arching co‐design viewpoint, this review assesses the use of cross‐bar based CIM for neural networks, connecting the material properties and the associated design constraints and demands to application, architecture, and performance. Both digital and analog memory are considered, assessing the status for training and inference, and providing metrics for the collective set of properties non‐volatile memory materials will need to demonstrate for a successful CIM technology.
There are many factors that affect the level of requirements volatility a system experiences over its lifecycle and the risk that volatility imparts. Improper requirements generation, undocumented user expectations, conflicting design decisions, and anticipated / unanticipated world states are representative of these volatility factors. Combined, these volatility factors can increase programmatic risk and adversely affect successful system development. This paper proposes that a Bayesian Network can be used to support reasonable judgments concerning the most likely sources and types of requirements volatility a developing system will experience prior to starting development and by doing so it is possible to predict the level of requirements volatility the system will experience over its lifecycle. This assessment offers valuable insight to the system's developers, particularly by providing a starting point for risk mitigation planning and execution.
Analog hardware accelerators, which perform computation within a dense memory array, have the potential to overcome the major bottlenecks faced by digital hardware for data-heavy workloads such as deep learning. Exploiting the intrinsic computational advantages of memory arrays, however, has proven to be challenging principally due to the overhead imposed by the peripheral circuitry and due to the non-ideal properties of memory devices that play the role of the synapse. We review the existing implementations of these accelerators for deep supervised learning, organizing our discussion around the different levels of the accelerator design hierarchy, with an emphasis on circuits and architecture. We explore and consolidate the various approaches that have been proposed to address the critical challenges faced by analog accelerators, for both neural network inference and training, and highlight the key design trade-offs underlying these techniques.
Nowadays, data generated by large-scale scientific experiments are on the scale of petabytes per month. These data are transferred through dedicated high-bandwidth networks across distributed collaboration sites for processing, storage, and analysis. Like general purpose networks, research networks experience attacks. However, monitoring traffic in such high-bandwidth network environments is challenging with conventional deep packet inspection (DPI)- based network intrusion detection systems (NIDS). Recently, machine learning (ML) algorithms have shown potential in detecting anomalous patterns with better generalizability and faster speed. Compared to the general purpose networks, traffic movement in research network is less volatile and more likely to benefit from the ML-based anomaly analysis. We present ThunderSecure, an efficient AI-powered system for detecting anomalies in 100G research networks. ThunderSecure implements a high-throughput packet processing infrastructure using multi-cores and GPUs. It takes time-series statistics from a variety of network measurements and uses 2D Convolutional Neural Network (CNN) to explore both spatial and temporal dependencies in network data stream. Patterns of science data traffic are learned with a one-class neural network, which blends an Adversarial Autoencoder (AAE) with Gaussian mixture density estimation. Testing traffic flows exhibiting significant deviations from the learned baseline of normality are marked as anomalies. We trained ThunderSecure on hundreds of billions of science data packets recorded from a 100G research network at Fermi National Accelerator Laboratory. The detection performance was evaluated on traffic captured from the same research network days and weeks after the training with different types of attack flows injected. To the best of our knowledge, this is the first ML work in the area of network anomaly detection that has been validated on such extreme scale datasets. Results show that ThunderSecure can recognize science data traffic that are captured long after the training and make nearly certain detection on those with anomalous flows injected, even in the case when the anomaly-to-normal mixing ratio is 0.1%.
With the expansion of renewable energy resources in the electric power systems, having flexibility in the setup will allow to maintain the system's reliability and prevailing operations. Such flexibility can be extracted from utility operated and/or consumer owned devices, such as, storage devices, electric vehicles, etc. For the demand side, generally consumer preferences, incentives, etc. enact on the availability of the flexibility; besides, both the spatial and temporal dimension dictates the degree of the flexibility. Consequently, the optimal dispatch of the grid resources might appear intractable as the considerable amount of flexibility are obliquely stemming from the ungovernable consumer devices. Thus characterizing the consequences of diverged feasible flexibility in the system is crucial for operations. In this paper, a procedure is developed to quantify the degree of flexibility of power systems in terms of resource dispatch reconfiguration. Specifically, we develop optimization problems to attain equivalent resource configurations for the power systems to evaluate the spatial volatility of the network and asses the flexibility of the system. The developed process is then validated using numerical simulations for IEEE-30 bus test system.
One of the major challenges for NASA's next-generation reusable-launch-vehicle (RLV) program is the design of a cryogenic lightweight composite fuel tank. Potential matrix resin systems need to exhibit a low coefficient of thermal expansion (CTE), good mechanical strength, and excellent barrier properties at cryogenic temperatures under load. In addition, the resin system needs to be processable by a variety of non-autoclavable techniques, such as vacuum-bag curing, resin-transfer molding (RTM), vacuum-assisted resin-transfer molding (VaRTM), resin-film infusion (RFI), pultrusion, and advanced tow placement (ATP). To meet these requirements, the Advanced Materials and Processing Branch (AMPB) at NASA Langley Research Center developed a new family of wholly aromatic liquid-crystal oligomers that can be processed and thermally cross-linked while maintaining their liquid-crystal order. All the monomers were polymerized in the presence of a cross-linkable unit by use of an environmentally benign melt-condensation technique. This method does not require hazardous solvents, and the only side product is acetic acid. The final product can be obtained as a powder or granulate and has an infinite shelf life. The obtained oligomers melt into a nematic phase and do not exhibit isotropization temperatures greater than the temperatures of decomposition (Ti > T(sub dec)). Three aromatic formulations were designed and tested and included esters, ester-amides, and ester-imides. One of the major advantages of this invention, named LaRC-LCR or Langley Research Center-Liquid Crystal Resin, is the ability to control a variety of resin characteristics, such as melting temperature, viscosity, and the cross-link density of the final part. Depending on the formulation, oligomers can be prepared with melt viscosities in the range of 10-10,000 poise (100 rad/s), which can easily be melt-processed using a variety of composite-processing techniques. This capability provides NASA with custom-made matrix resins that meet the required processing conditions for the fabrication of textile composites. Once the resin is in place, the temperature is raised to 375 C and the oligomers are cross-linked into a high-glass-transition-temperature (Tg) nematic network without releasing volatiles. The mechanical properties of the fully crosslinked, composite articles are comparable to typical composites based on commercially available epoxy resins.
The Resource Prospector (RP) mission with the Regolith and Environment Science and Oxygen Lunar Volatile Extraction (RESOLVE) payload aims to show the presence of water in lunar regolith, and establish a proving ground for NASAs mission to Mars. One of the analysis is performed by the Lunar Advanced Volatiles Analysis (LAVA) subsystem, which consists of a fluid network that facilitates the transport of volatile samples to a gas chromatograph and mass spectrometer (GC-MS) instrument. The understanding of fluid dynamics directed from the GC to the MS is important due to the influence of flow rates and pressures that affect the accuracy of and prevent the damage to the overall GC-MS instrument. The micro-scale capillary fluid network within the GC alone has various lengths and inner-diameters; therefore, determination of pressure differentials and flow rates are difficult to model computationally, with additional complexity from the vacuum conditions in space and lack of a lunar atmosphere. A series of tests were performed on an experimental set-up of the system where the inner diameters of the GC transfer line connecting to the MS were varied. The effect on chromatography readings were also studied by applying these lines onto a GC instrument. It was found that a smaller inner diameter transfer line resulted in a lower flow rate, as well as a lower pressure differential across the thermal conductivity detector (TCD) unit of the GC and a negligible pressure drop across the mock-up capillary column. The chromatography was affected with longer retention times and broader peak integrations. It was concluded that a 0.050 mm inner diameter line still proved most suitable for the systems flow rate preferences. In addition, it was evident that this small transfer line portrayed some expense to GC signal characteristics and the wait time for steady-state operation.
High penetration from volatile renewable energy resources in the grid and the varying nature of loads raise the need for frequent line switching to ensure the efficient operation of electrical distribution networks. Operators must ensure maximum load delivery, reduced losses, and the operation between voltage limits. However, computations to decide the optimal feeder configuration are often computationally expensive and intractable, making it unfavorable for real-time operations. This is mainly due to the existence of binary variables in the network reconfiguration optimization problem. To tackle this issue, we have devised an approach that leverages machine learning techniques to reshape distribution networks featuring multiple substations. This involves predicting the substation responsible for serving each part of the network. Hence, it leaves simple and more tractable Optimal Power Flow problems to be solved. This method can produce accurate results in a significantly faster time, as demonstrated using the IEEE 37-bus distribution feeder. Compared to the traditional optimization-based approaches, a feasible solution is achieved approximately ten times faster for all the tested scenarios.
The operation of distribution networks is becoming increasingly volatile, due to fast variations of renewables and, hence, net-loading conditions. To perform a reliable state estimation under these conditions, this paper considers the case where measurements from meters, phasor measurement units, and distributed energy resources are collected and processed in real time to produce estimates of the state at a fast time scale. Streams of measurements collected in real time and at heterogenous rates render the underlying processing asynchronous, and poses severe strains on workhorse state estimation algorithms. In this work, a real-time state estimation algorithm is proposed, where data are processed on the fly. Starting from a regularized least-squares model, and leveraging appropriate linear models, the proposed scheme boils down to a linear dynamical system where the state is updated based on the previous estimate and on the measurement gathered from a few available sensors. The estimation error is shown to be always bounded under mild condition. Numerical simulations are provided to corroborate the analytical findings.
Exploiting the prominent role of complex memories in exascale node architecture, the UMap page fault handler offers new capabilities to access large memory-mapped data sets directly. UMap provides flexible configuration options to customize page handling to each application, including analysis of massive observational and simulation data sets. The high-performance design features I/O decoupling, dynamic load balancing, and application-level controls. Page faults triggered by application threads and processes accessing data mapped to a UMapp’ed region are handled via the Linux userfaultfd protocol, an asynchronous message-oriented kernel-user communication mechanism that avoids the context switch penalty of traditional signal fault handlers. UMap is fully open source. In this paper, we give an overview of the UMap library architecture, its extensible plugin architecture, and the use/performance of UMap in emerging heterogeneous memory hierarchies such as near-node Non-volatile Memory (NVM) and network attached memories. We highlight new capabilities in two pagefault management plugins, the NetworkStore and SparseStore. We demonstrate the integration between UMap and multiple ECP products including Caliper, Metall, ZFP, Mochi, and Ripples.
The operation of distribution networks is becoming increasingly volatile, due to fast variations of renewables and, hence, net-loading conditions. To perform a reliable state estimation under these conditions, this paper considers the case where measurements from meters, phasor measurement units, and distributed energy resources are collected and processed in real time to produce estimates of the state at a fast time scale. Streams of measurements collected in real time and at heterogenous rates render the underlying processing asynchronous, and poses severe strains on workhorse state estimation algorithms. In this work, a real-time state estimation algorithm is proposed, where data are processed on the fly. Starting from a regularized least-squares model, and leveraging appropriate linear models, the proposed scheme boils down to a linear dynamical system where the state is updated based on the previous estimate and on the measurement gathered from a few available sensors. The estimation error is shown to be always bounded under mild condition. Numerical simulations are provided to corroborate the analytical findings.
Secondary organic aerosols (SOA) are large contributors to fine particle loadings and radiative forcing, but are often represented crudely in global models. We have implemented three new detailed SOA treatments within the Community Atmosphere Model version 5 (CAM5) that allow us to compare the semi-volatile versus non-volatile SOA treatments (based on some of the latest experimental findings) and also investigate the effects of gas-phase fragmentation reactions. For semi-volatile SOA treatments, fragmentation reactions decrease simulated SOA burden from 7.5 Tg to 1.8 Tg. For the non-volatile SOA treatment with fragmentation, the burden is 3.1 Tg. Larger differences between non-volatile and semi-volatile SOA (upto a factor of 5) correspond to continental outflow over the oceans. Compared to a global dataset of surface Aerosol Mass Spectrometer measurements and the US IMPROVE network measurements, the non-volatile SOA with fragmentation treatment (FragNVSOA) agrees best at rural locations. Urban SOA is under-predicted but this may be due to the coarse model resolution. All our three revised treatments show much better agreement with aircraft measurements of organic aerosols (OA) over the N. American Arctic and sub-Arctic in spring and summer, compared to the standard CAM5 formulation. This is due to treating SOA precursor gases from biomass burning, and long-range transport of biomass burning OA at elevated levels. The revised model configuration that include fragmentation (both semi-volatile and non-volatile SOA) show much better agreement with MODIS AOD data over regions dominated by biomass burning during the summer, and predict biomass burning as the largest global source of OA followed by biogenic and anthropogenic sources. The non-volatile and semi-volatile configuration predict the direct radiative forcing of SOA as -0.5 W m-2 and -0.26 W m-2 respectively, at top of the atmosphere, which are higher than previously estimated by most models, but in reasonable agreement with a recent constrained modeling study. This study highlights the importance of improving process-level representation of SOA in global models.
Ion Mobility Spectrometry (IMS) is a powerful technique for trace organic analysis in the gas phase. Quantitative measurements are difficult, because IMS has a limited linear range. Factors that may affect the instrument response are pressure, temperature, and humidity. Nonlinear calibration methods, such as neural networks, may be ideally suited for IMS. Neural networks have the capability of modeling complex systems. Many neural networks suffer from long training times and overfitting. Cascade correlation neural networks train at very fast rates. They also build their own topology, that is a number of layers and number of units in each layer. By controlling the decay parameter in training neural networks, reproducible and general models may be obtained.
A new experimental approach is presented in which two separate cryogenic ion traps are used to reproducibly form weakly bound solvent clusters around electrosprayed ions and messenger-tag them for single-photon infrared photodissociation spectroscopy. This approach thus enables the vibrational characterization of ionic clusters comprised of a solvent network around large and non-volatile ions. We demonstrate the capabilities of the instrument by clustering water, methanol, and acetone around a protonated glycylglycine peptide. For water, cluster sizes with greater than twenty solvent molecules around a single ion are readily formed. We further demonstrate that similar water clusters can be formed around ions having a shielded charge center or those that do not readily form hydrogen bonds. Finally, infrared photodissociation spectra of D{sub 2}-tagged GlyGlyH{sup +} ⋅ (H{sub 2}O){sub 1−4} are presented. They display well-resolved spectral features and comparisons with calculations reveal detailed information on the solvation structures of this prototypical peptide.
Sandia is researching novel methods of analysis for large datasets. The term “Big Data” has been used frequently in recent years to describe a discipline of computer science that attempts to distil an increasing quantity of data into a result or meaningful subset that can be easily visualized and acted upon by human consumers. These data are frequently characterized by one or more of the following: high volume, velocity, variety, and veracity. Sandia scientists and researchers are no strangers to large datasets, which often typify the result of high-performance computing workloads. The challenge for Sandia data scientists, however, comes at the arrival of exascale computing, which has become mandated by executive order for national computing centers. Because of power and networking limitations, exascale computing precludes the movement of datasets at exascale from data production to data analysis. Such analysis must be done in-situ, to mitigate the time and power costs of network transport. Currently, many of Sandia’s data analytics environments exist as independent entities, often specialized for individual use cases, interconnected to other data analytics, high-performance computing, and storage systems. Sandia is procuring, installing, and researching future platforms for such data analytics. A memory footprint of 256 GB per node, fourteen data rate (FDR) InfiniBand networking, PCIe connected non-volatile memory (NVMe) data storage, and an object data store will characterize the coming institutional data analytics platforms. The combined systems will contain 300 compute nodes. A cloud-inspired software stack resulting from collaboration between the Sandia centers will allow the systems to support a flexible mix of user and system-facing workloads and use cases including code development and testing in R and Python, Apache Hadoop and Spark, distributed in-memory databases, graph analytics, emulytics, system log and predictive analytics, and data visualization for institutional and mission customers. These systems will arrive at the end of FY 2015 and production is planned by the spring of FY 2016.