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 127 records · Page 7

GeTe sequences in superlattice phase change memories and their electrical characteristics

We studied GeTe structures in superlattice phase change memories (superlattice PCMs) with a [GeTe/Sb{sub 2}Te{sub 3}] stacked structure by X-ray diffraction (XRD) analysis. We examined the electrical characteristics of superlattice PCMs with films deposited at different temperatures. It was found that XRD spectra differed between the films deposited at 200 °C and 240 °C; the differences corresponded to the differences in the GeTe sequences in the films. We applied first-principles calculations to calculate the total energy of three different GeTe sequences. The results showed the Ge-Te-Ge-Te sequence had the lowest total energy of the three and it was found that with this sequence the superlattice PCMs did not run.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗

Zero-Copy Objects System

Zero-Copy Objects System software enables application data to be encapsulated in layers of communication protocol without being copied. Indirect referencing enables application source data, either in memory or in a file, to be encapsulated in place within an unlimited number of protocol headers and/or trailers. Zero-copy objects (ZCOs) are abstract data access representations designed to minimize I/O (input/output) in the encapsulation of application source data within one or more layers of communication protocol structure. They are constructed within the heap space of a Simple Data Recorder (SDR) data store to which all participating layers of the stack must have access. Each ZCO contains general information enabling access to the core source data object (an item of application data), together with (a) a linked list of zero or more specific extents that reference portions of this source data object, and (b) linked lists of protocol header and trailer capsules. The concatenation of the headers (in ascending stack sequence), the source data object extents, and the trailers (in descending stack sequence) constitute the transmitted data object constructed from the ZCO. This scheme enables a source data object to be encapsulated in a succession of protocol layers without ever having to be copied from a buffer at one layer of the protocol stack to an encapsulating buffer at a lower layer of the stack. For large source data objects, the savings in copy time and reduction in memory consumption may be considerable.

Burleigh, Scott C.↗

Effect of different stages of deformation on the microstructure evolution of Ti-rich NiTi shape memory alloy

The main objective of this work was to investigate the thermomechanical behavior and microstructural changes of a Ti-rich NiTi shape memory alloy (SMA). The microstructural and texture evolution of aged NiTi alloy at different degrees of deformation were elicited by transmission electron microscopy (TEM). An effort was made to correlate results obtained from the tensile test with results from microstructure studies. The undeformed sample reveals a self-accommodated morphology with straight and well defined twin boundaries. At different stages of deformation, diverse mechanisms were involved. These mechanisms include marstraining, detwinning accompanied by dislocation movement, and finally, severe plastic deformation, subdivision and amorphization of the matrix. Under increasing strains, high density lattice defects were generated and the morphology of B19’ became disordered. - Graphical abstract: The summary of microstructure changes of the martensite twins during tensile deformation in polycrystalline NiTi SMAs. - Highlights: • Initial elastic response, dislocation avalanche and deformation bands were studied. • < 011 > Type II twin accompanied by detwinned area after 2% cold work was observed. • Visible parallel fine stacking faults showed plastic flow of the material. • At higher strains, subgrains changed to recrystallized, finely amorphous structure.

36 MATERIALS SCIENCE↗

Packaging of a large capacity magnetic bubble domain spacecraft recorder

A Solid State Spacecraft Data Recorder (SSDR), based on bubble domain technology, having a storage capacity of 10 to the 8th power bits, was designed and is being tested. The recorder consists of two memory modules each having 32 cells, each cell containing sixteen 100 kilobit serial bubble memory chips. The memory modules are interconnected to a Drive and Control Unit (DCU) module containing four microprocessors, 500 integrated circuits, a RAM core memory and two PROM's. The two memory modules and DCU are housed in individual machined aluminum frames, are stacked in brick fashion and through bolted to a base plate assembly which also houses the power supply.

Becker, F. J.↗

Thermal conductivity measurement of amorphous dielectric multilayers for phase-change memory power reduction

In this work, we investigate the temperature-dependent thermal conductivities of few nanometer thick alternating stacks of amorphous dielectrics, specifically SiO{sub 2}/Al{sub 2}O{sub 3} and SiO{sub 2}/Si{sub 3}N{sub 4}. Experiments using steady-state Joule-heating and electrical thermometry, while using a micro-miniature refrigerator over a wide temperature range (100–500 K), show that amorphous thin-film multilayer SiO{sub 2}/Si{sub 3}N{sub 4} and SiO{sub 2}/Al{sub 2}O{sub 3} exhibit through-plane room temperature effective thermal conductivities of about 1.14 and 0.48 W/(m × K), respectively. In the case of SiO{sub 2}/Al{sub 2}O{sub 3}, the reduced conductivity is attributed to lowered film density (7.03 → 5.44 × 10{sup 28 }m{sup –3} for SiO{sub 2} and 10.2 → 8.27 × 10{sup 28 }m{sup –3} for Al{sub 2}O{sub 3}) caused by atomic layer deposition of thin-films as well as a small, finite, and repeating thermal boundary resistance (TBR) of 1.5 m{sup 2} K/GW between dielectric layers. Molecular dynamics simulations reveal that vibrational mismatch between amorphous oxide layers is small, and that the TBR between layers is largely due to imperfect interfaces. Finally, the impact of using this multilayer dielectric in a dash-type phase-change memory device is studied using finite-element simulations.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Interconnect architecture for three-dimensional processing systems

A processing system includes a plurality of processor cores formed in a first layer of an integrated circuit device and a plurality of partitions of memory formed in one or more second layers of the integrated circuit device. The one or more second layers are deployed in a stacked configuration with the first layer. Each of the partitions is associated with a subset of the processor cores that have overlapping footprints with the partitions. The processing system also includes first memory paths between the processor cores and their corresponding subsets of partitions. The processing system further includes second memory paths between the processor cores and the partitions.

Jayasena, Nuwan S.↗

Exploring Architectural-Aware Affinity Policies in Modern HPC Runtimes

Modern commodity and High-Performance Computing (HPC) systems are evolving with complex CPU architectures. These architectures now feature higher core and NUMA domain counts and implement features such as hyperthreading. When considering significant differences in hardware configurations, library availability, and hardware-tailored system/software stacks, which could substantially vary from one system to another, performance portability is hard to achieve. Throughout the years, this trend resulted in an increasingly high burden on application developers to fine-tune their workloads for each architecture. This work explores how hardware-dependent aspects such as locality/process/thread affinity affect performance in modern CPU architectures. We focus our study on the Global Memory and Threading (GMT) distributed runtime system as a representative of Partitioned Global Address Space (PGAS) software stacks commonly adopted for productivity. In particular, to appreciate performance implications, we evaluate GMT’s thread affinity policies, and, introduce two new ones which exploit architectural awareness. Finally, we explore alternative NUMA configurations via different process bindings and perform a scalability study on three HPC clusters with varying CPU architectures and NUMA layouts. Our analysis indicates that more complex architectures are more affected by affinity and binding policies and highlights the importance of setting proper runtime configurations to achieve superior performance.

Di Dio Lavore, Ian↗

Sustainable Shape Memory Elastomers with Reduced Melt Viscosity and Enhanced Stiffness

Melt reactive processing of lignin with nitrile rubber is a promising approach to synthesizing shape memory materials. The strong intramolecular interactions in lignin macromolecular structures, caused by π–π stacking in aromatic rings and hydrogen bonding, often result in large phase separation or low miscibility with rubbers. In this study, we investigated the chemical and molecular characteristics, as well as the stiffness and complex viscosity, of modified kraft lignin melt-reacted with an acrylonitrile/butadiene copolymer containing 41% acrylonitrile (NBR41). To enhance the macromolecular compatibility of kraft lignin with NBR41, kraft lignin was cross-linked with poly(propylene glycol) diglycidyl ether (PPDE) and trimethylolpropane triglycidyl ether (TTE), both rich in epoxy reactive groups capable of forming chemical bonds with hydroxyl and carboxyl groups. Here, our findings demonstrate that the modification of kraft lignin with PPDE and TTE resulted in significantly increased stiffness of the composites. The elastic modulus of NBR41-Kraft lignin-PPDE and NBR41-Kraft lignin-TTE increased by 82 and 162%, respectively. Both the yield strength and Young’s modulus of these two samples showed dramatic improvements. Specifically, the yield strength and Young’s modulus of NBR41-Kraft lignin-TTE increased nearly 4 and 3-fold, respectively, compared to the control sample. Interestingly, despite significant improvements in mechanical properties, the viscosity of NBR41-Kraft lignin-PPDE was substantially lower than that of the control sample. At 210 °C and an angular frequency of 1 rad/s, the complex viscosity of NBR41-Kraft lignin was approximately 100.25 ± 4.77 kPa·s, while that of NBR41-Kraft lignin-PPDE was significantly lower at 56 ± 0.93 kPa·s. These findings were validated through Fourier transform infrared spectroscopy, scanning electron microscopy, dynamic mechanical analysis, thermal characterization, rheological tests, and quasi-elastic neutron scattering techniques.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

A uGNI-Based Asynchronous Message-driven Runtime System for Cray Supercomputers with Gemini Interconnect

Gemini as the network for new Cray XE/XT systems features low latency, high bandwidth and strong scalability. Its hardware support for remote direct memory access enables efficient implementation of the global address space programming languages. Although the Generic Network Interface (GNI) is designed to support message-passing applications, it is still challenging to attain good performance for applications written in alternative programming models, such as the message-driven programming model. In our earlier work we showed that CHARM++, an object-oriented message-driven programming model, scales up to the full Jaguar Cray machine. In this paper, we describe a general and light-weight asynchronous Low-level RunTime System (LRTS) for CHARM+, and its implementation on the uGNI software stack for Cray XE systems. Several techniques are presented to exploit the uGNI capability by reducing memory copy and registration overhead, taking advantage of persistent communication, and improving intra-node communication. Our micro-benchmark results demonstrate that the uGNI-based runtime system outperforms the MPI-based implementation by up to 50% in terms of message latency. For communication intensive applications such as N-Queens, this implementation scales up to 15,360 cores of a Cray XE6 machine and is 70% faster than an MPI-based implementation. In molecular dynamics application NAMD, the performance is also considerably improved by as high as 18%.

Asynchronous message-passing↗

Opportunities for nonvolatile memory systems in extreme-scale high-performance computing

For extreme-scale high-performance computing systems, system-wide power consumption has been identified as one of the key constraints moving forward, where DRAM main memory systems account for about 30 to 50 percent of a node's overall power consumption. As the benefits of device scaling for DRAM memory slow, it will become increasingly difficult to keep memory capacities balanced with increasing computational rates offered by next-generation processors. However, several emerging memory technologies related to nonvolatile memory (NVM) devices are being investigated as an alternative for DRAM. Moving forward, NVM devices could offer solutions for HPC architectures. Researchers are investigating how to integrate these emerging technologies into future extreme-scale HPC systems and how to expose these capabilities in the software stack and applications. In addition, current results show several of these strategies could offer high-bandwidth I/O, larger main memory capacities, persistent data structures, and new approaches for application resilience and output postprocessing, such as transaction-based incremental checkpointing and in situ visualization, respectively.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗

Lunar Passive Fuel Cell Technologies: Evaporator + Fluid Polishing Elements

This project investigates key elements of a fully-passive, fuel-cell based, surface power generation concept. The overall concept replaces the actively pumped thermal management of a state of the art (SOA) fuel cell/electrolyzer stack with a passive 2-phase thermosyphon for heat transport and a passive shape memory actuating radiator for temperature management. The elements of the concept investigated in this project are: Evaporator Element: Evaluating the compatibility of common fuel-cell interconnect materials in the evaporator of a 2-phase water thermosyphon Fluid Polishing element: The use of membrane gas separation and resin water polishing to address non-condensable gas removal and water purity.

Fuel Cell↗

Passive fuel-Cell Surface power System (PaCeSS)

This effort produces a fully-passive surface power generation capability for fuel cell technology. The concept replaces the actively pumped thermal management of a state of the art (SOA) fuel cell/electrolyzer stack with a passive 2-phase thermosyphon for heat transport and a passive shape memory actuating radiator for temperature management.

Fuel Cell↗

Tuning resistance states by thickness control in an electroforming-free nanometallic complementary resistance random access memory

Tuning low resistance state is crucial for resistance random access memory (RRAM) that aims to achieve optimal read margin and design flexibility. By back-to-back stacking two nanometallic bipolar RRAMs with different thickness into a complementary structure, we have found that its low resistance can be reliably tuned over several orders of magnitude. Such high tunability originates from the exponential thickness dependence of the high resistance state of nanometallic RRAM, in which electron wave localization in a random network gives rise to the unique scaling behavior. The complementary nanometallic RRAM provides electroforming-free, multi-resistance-state, sub-100 ns switching capability with advantageous characteristics for memory arrays.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

A Parallel Graph Environment for Real-World Data Analytics Workflows

Economic competitiveness and national security depend increasingly on the insightful analysis of large data sets. The diversity of real-world data sources and analytic workflows impose challenging hardware and software requirements for parallel graph platforms. The irregular nature of graph methods is not supported well by the deep memory hierarchies of conventional distributed systems, requiring new processor and runtime system designs to tolerate memory and synchronization latencies. Moreover, the efficiency of relational table operations and matrix computations are not attainable when data is stored in common graph data structures. In this paper, we present HAGGLE, a high-performance, scalable data analytics platform. The platform’s hybrid data model supports a variety of distributed, thread-safe data structures, parallel programming constructs, and persistent and streaming data. An abstract runtime layer enables us to map the stack to conventional, distributed computer systems with accelerators. The runtime uses multithreading, active messages, and data aggregation to hide memory and synchronization latencies on large-scale systems.

Graph Analytics↗

Cache-based error recovery for shared memory multiprocessor systems

A multiprocessor cache-based checkpointing and recovery scheme for of recovering from transient processor errors in a shared-memory multiprocessor with private caches is presented. New implementation techniques that use checkpoint identifiers and recovery stacks to reduce performance degradation in processor utilization during normal execution are examined. This cache-based checkpointing technique prevents rollback propagation, provides for rapid recovery, and can be integrated into standard cache coherence protocols. An analytical model is used to estimate the relative performance of the scheme during normal execution. Extensions that take error latency into account are presented.

Wu, Kun-Lung↗

Weaves as an Interconnection Fabric for ASIM's and Nanosatellites

Many of the micromachines under consideration require computer support, indeed, one of the appeals of this technology is the ability to intermix mechanical, optical, analog, and digital devices on the same substrate. The amount of computer power is rarely an issue, the sticking point is the complexity of the software required to make effective use of these devices. Micromachines are the nano-technologist's equivalent of 'golden screws'. In other words, they will be piece parts in larger assemblages. For example, a nano-satellite may be composed of stacked silicon wafers where each wafer contains hundreds to thousands of micromachines, digital controllers, general purpose computers, memories, and high-speed bus interconnects. Comparatively few of these devices will be custom designed, most will be stock parts selected from libraries and catalogs. The novelty will lie in the interconnections. For example, a digital accelerometer may be a component part in an adaptive suspension, a monitoring element embedded in the wrapper of a package, or a portion of the smart skin of a launch vehicle. In each case, this device must inter-operate with other devices and probes for the purposes of command, control, and communication. We propose a software technology called 'weaves' that will permit large collections of micromachines and their attendant computers to freely intercommunicate while preserving modularity, transparency, and flexibility. Weaves are composed of networks of communicating software components. The network, and the components comprising it, may be changed even while the software, and the devices it controls, are executing. This unusual degree of software plasticity permits micromachines to dynamically adapt the software to changing conditions and allows system engineers to rapidly and inexpensively develop special purpose software by assembling stock software components in custom configurations.

Gorlick, Michael M.↗

The Design and Implementation of NASA's Advanced Flight Computing Module

This paper describes a working flight computer Multichip Module developed jointly by JPL and TRW under their respective research programs in a collaborative fashion. The MCM is fabricated by nCHIP and is packaged within a 2 by 4 inch Al package from Coors. This flight computer module is one of three modules under development by NASA's Advanced Flight Computer (AFC) program. Further development of the Mass Memory and the programmable I/O MCM modules will follow. The three building block modules will then be stacked into a 3D MCM configuration. The mass and volume of the flight computer MCM achieved at 89 grams and 1.5 cubic inches respectively, represent a major enabling technology for future deep space as well as commercial remote sensing applications.

computer design space applications multichip modul↗

High Performance Data Analytics on a Commodity Private Cloud

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.

97 MATHEMATICS AND COMPUTING↗