Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “computer hardware”

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 1,063 records · Page 59

Impact of Emerging Computing Architectures and Opportunities for Process Systems Engineering Applications

Moore’s “law” was the observation that the number of transistors in an integrated circuit doubled approximately every two years. This trend has distinctly failed to hold in recent years. The death of Moore’s law has left researchers and practitioners in the computational sciences searching for technologies to provide the speedups formerly supported by Moore’s law. Previously overlooked chip architectures and other computing technologies are now receiving more development resources. Critically, these technologies are gaining more mature software support, opening their adoption by researchers in algorithms and applications. In this article, we review some of these computing technologies, their relationship with various algorithms and applications, and their potential benefits (or pitfalls). We close with recommendations for future work by the process systems engineering community specifically.

Emerging hardware↗

Augmentation of the space station module power management and distribution breadboard

The space station module power management and distribution (SSM/PMAD) breadboard models power distribution and management, including scheduling, load prioritization, and a fault detection, identification, and recovery (FDIR) system within a Space Station Freedom habitation or laboratory module. This 120 VDC system is capable of distributing up to 30 kW of power among more than 25 loads. In addition to the power distribution hardware, the system includes computer control through a hierarchy of processes. The lowest level consists of fast, simple (from a computing standpoint) switchgear that is capable of quickly safing the system. At the next level are local load center processors, (LLP's) which execute load scheduling, perform redundant switching, and shed loads which use more than scheduled power. Above the LLP's are three cooperating artificial intelligence (AI) systems which manage load prioritizations, load scheduling, load shedding, and fault recovery and management. Recent upgrades to hardware and modifications to software at both the LLP and AI system levels promise a drastic increase in speed, a significant increase in functionality and reliability, and potential for further examination of advanced automation techniques. The background, SSM/PMAD, interface to the Lewis Research Center test bed, the large autonomous spacecraft electrical power system, and future plans are discussed.

Walls, Bryan↗

Simulating strongly correlated molecules with a superconducting quantum processor

Many of the biggest challenges in expanding the nation’s access to clean and low-cost energy resources are fundamentally chemistry or materials challenges. An important case is the development of new catalysts for the up-conversion of cheap and readily available materials such as methane or water into materials suitable for use as a fuel such as methanol or oxygen. To understand and exploit such processes, computer simulations of chemical reactions provide a natural complement to experimental studies. Unfortunately, most catalytic reactions involve so-called “strongly correlated” molecules which are notoriously difficult to study with simulation algorithms that can be executed on existing (classical) computers. The recent growth in quantum information science offers an alternative potential route for simulating these difficult systems. As a result, an increasing number of computational chemists are becoming interested in quantum computing. At the same time, quantum information scientists have identified chemistry simulation as a possible first demonstration of a quantum computer providing an improvement over a classical computer. The objective of this project is to accurately simulate strongly correlated molecules on a quantum processor. To meet the high challenges of this objective, new hybrid quantum/classical algorithms will be co-designed with advanced quantum gate developments and computed on customized quantum hardware. Some of the developed techniques will be transferable to study other molecular systems, while the project as a whole will help define better strategies for advancing the quantum simulation of matter more generally.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Machine Learning-enabled Scalable Performance Prediction of Scientific Codes

Hardware architectures become increasingly complex as the compute capabilities grow to exascale. Here, we present the Analytical Memory Model with Pipelines (AMMP) of the Performance Prediction Toolkit (PPT). PPT-AMMP takes high-level source code and hardware architecture parameters as input and predicts runtime of that code on the target hardware platform, which is defined in the input parameters. PPT-AMMP transforms the code to an (architecture-independent) intermediate representation, then (i) analyzes the basic block structure of the code, (ii) processes architecture-independent virtual memory access patterns that it uses to build memory reuse distance distribution models for each basic block, and (iii) runs detailed basic-block level simulations to determine hardware pipeline usage. PPT-AMMP uses machine learning and regression techniques to build the prediction models based on small instances of the input code, then integrates into a higher-order discrete-event simulation model of PPT running on Simian PDES engine. We validate PPT-AMMP on four standard computational physics benchmarks and present a use case of hardware parameter sensitivity analysis to identify bottleneck hardware resources on different code inputs. We further extend PPT-AMMP to predict the performance of a scientific application code, namely, the radiation transport mini-app SNAP. To this end, we analyze multi-variate regression models that accurately predict the reuse profiles and the basic block counts. We validate predicted SNAP runtimes against actual measured times.

97 MATHEMATICS AND COMPUTING↗

Quantum Computer-Aided Design: Digital Quantum Simulation of Quantum Processors

With the increasing size of quantum processors, submodules that constitute the processor hardware will become too large to accurately simulate on a classical computer. Therefore, one would soon have to fabricate and test each new design primitive and parameter choice in time-consuming coordination between design, fabrication, and experimental validation. Here we show how one can design and test the performance of next-generation quantum hardware—by using existing quantum computers. Focusing on superconducting transmon processors as a prominent hardware platform, we compute the static and dynamic properties of individual and coupled transmons. We show how the energy spectra of transmons can be obtained by variational hybrid quantum-classical algorithms that are well suited for near-term noisy quantum computers. In addition, single- and two-qubit gate simulations are demonstrated via Suzuki-Trotter decomposition. Our methods pave a promising way towards designing candidate quantum processors when the demands of calculating submodule properties exceed the capabilities of classical computing resources.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Applications of NMR and computational methodologies to study protein dynamics

The overwhelming evidence now illustrates the defining role of atomic-scale protein flexibility in biological events such as allostery, cell signaling, and enzyme catalysis. Over the years, spin relaxation nuclear magnetic resonance (NMR) has provided significant insights on the structural motions occurring on multiple time frames over the course of a protein life span. The present review article aims to illustrate to the broader community how this technique continues to shape many areas of protein science and engineering, in addition to being an indispensable tool for studying atomic-scale motions and functional characterization. Continuing developments in underlying NMR technology alongside software and hardware developments for complementary computational approaches now enable methodologies to routinely provide spatial directionality and structural representations traditionally harder to achieve solely using NMR spectroscopy. In addition to its well-established role in structural elucidation, we present recent examples that illustrate the combined power of selective isotope labeling, relaxation dispersion experiments, chemical shift analyses, and computational approaches for the characterization of conformational sub-states in proteins and enzymes.

59 BASIC BIOLOGICAL SCIENCES↗

Survivable algorithms and redundancy management in NASA's distributed computing systems

The design of survivable algorithms requires a solid foundation for executing them. While hardware techniques for fault-tolerant computing are relatively well understood, fault-tolerant operating systems, as well as fault-tolerant applications (survivable algorithms), are, by contrast, little understood, and much more work in this field is required. We outline some of our work that contributes to the foundation of ultrareliable operating systems and fault-tolerant algorithm design. We introduce our consensus-based framework for fault-tolerant system design. This is followed by a description of a hierarchical partitioning method for efficient consensus. A scheduler for redundancy management is introduced, and application-specific fault tolerance is described. We give an overview of our hybrid algorithm technique, which is an alternative to the formal approach given.

Malek, Miroslaw↗

Hardware Evaluation: Abstract Machine Models and Proxy Architectures for Exascale Computing

The strategic and economic impacts of high performance computing (HPC) systems, over the last several decades, have enabled dramatic improvements in manufacturing, design, development, and research, across almost every sector of the U.S. economy, including genetic analysis for agriculture and medicine, oil exploration, electronics, aircraft, and automotive design, and of course national defense technologies. HPC has repeatedly compounded reductions in time-to-solution and time-to-market in each of these areas. In recognition of these strategic and economic impacts, the Department of Energy (DOE) created the Exascale Computing Project (ECP) as a broad-based technology project focusing on simultaneously co-evolving high-performance computing architecture, system software, and application software. While many details of the exascale architectures are undefined, communicating the form and direction of this co-evolution of advances, in each of these three components, across the project, presents a need for a common language through which these advances can be shared. In this document we present a series of abstract representations designed to allow application developers to focus on the aspects of the machine that are important or relevant to performance and code structure. These abstract machine models (AMMs) describe the proposed architectures at the component and system level and are intended as communication aids between application developers and hardware architects during the co-design process. In addition to providing a common language for communication, the constraints of protecting the intellectual property of ECP’s vendor partners requires that the architectural advances be presented in an abstract sense, rather than exposing the specifics of an individual vendor’s designs, while still providing enough detail such that application developers will be able to reason about the performance trade-offs in the design space of their applications.

97 MATHEMATICS AND COMPUTING↗

Three-Dimensional Integration of Functional Oxides and Crystalline Silicon for Optical Neuromorphic Computing Using Nanometer-Scale Oxygen Scavenging Barriers

Dating to the first reports of epitaxial oxide deposition on Si(001), the integration of complex oxides on silicon has been a fast-moving area of research, where fundamental materials physics is intimately connected to tremendous technological promise in such areas as integrated electronics, optical neuromorphic and quantum computing, and sensing, to name a few. Despite their great promise, devices relying on the co-integration of silicon and epitaxial perovskites are typically limited to basic, planar geometries due to practical issues with their fabrication. In this paper, we overcome these long-standing challenges by developing a method to produce high-quality Si(001)/TMO/Si(001)/TMO heterostructures without wafer bonding, resulting in the straightforward three-dimensional integration of functional complex oxides and active Si(001) layers into a technologically relevant platform that is needed for on-chip hardware implementations of neuromorphic computing based on optical signals. We present detailed structural and chemical characterization of our heterostructures and discuss generalized design rules for their fabrication. Our results exponentially expand the universe of practically achievable TMO-based integrated devices and push this promising class of materials closer to realizing their full technological potential.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

An FPGA-Based Neuromorphic Processor with All-to-All Connectivity

Neuromorphic computing is a promising paradigm for future energy-efficient computing. At present, however, it is in its nascent stages—most hardware implementations are research-grade, commercial products are not available, and the software tools are not production-ready. The lack of hardware and software tools makes neuromorphic computing inaccessible to researchers around the globe. To this extent, we intend to build a low-cost, open-source, FPGA-based digital neuromorphic processor that can be used by researchers worldwide. In this paper, we present a preliminary implementation of the processor on a Xilinx Artix-7 FPGA using SystemVerilog. Our implementation supports the integrate-and-fire neuron with two parameters each for neurons and synapses. It also features all-to-all connectivity among neurons on the hardware. We test our implementation on four cases: bars and stripes datasets, shortest path algorithm, logic gates, and 8-3 encoder. We also perform a scalability study to understand the resource utilization of the FPGA as the number of all-to-all connected neurons increases. With our implementation, the Artix- 7 supports 65 neurons with all-to-all connectivity. Moreover, all the test cases mentioned above achieve 100% accuracy.

Maheshwari, Disha↗

Quantum computation of frequency-domain molecular response properties using a three-qubit iToffoli gate

The quantum computation of molecular response properties on near-term quantum hardware is a topic of substantial interest. Computing these properties directly in the frequency domain is desirable, but the circuits require large depth if the typical hardware gate set consisting of single- and two-qubit gates is used. While high-fidelity multipartite gates have been reported recently, their integration into quantum simulation and the demonstration of improved accuracy of the observable properties remains to be shown. Here, we report the application of a high-fidelity multipartite gate, the iToffoli gate, to the computation of frequency-domain response properties of diatomic molecules. The iToffoli gate enables a ~50% reduction in circuit depth and ~40% reduction in circuit execution time compared to the traditional gate set. We show that the molecular properties obtained with the iToffoli gate exhibit comparable or better agreement with theory than those obtained with the native CZ gates. Our work is among the first demonstrations of the practical usage of a native multi-qubit gate in quantum simulation, with diverse potential applications to near-term quantum computation.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Results of solar electric thrust vector control system design, development and tests

Efforts to develop and test a thrust vector control system TVCS for a solar-energy-powered ion engine array are described. The results of solar electric propulsion system technology (SEPST) III real-time tests of present versions of TVCS hardware in combination with computer-simulated attitude dynamics of a solar electric multi-mission spacecraft (SEMMS) Phase A-type spacecraft configuration are summarized. Work on an improved solar electric TVCS, based on the use of a state estimator, is described. SEPST III tests of TVCS hardware have generally proved successful and dynamic response of the system is close to predictions. It appears that, if TVCS electronic hardware can be effectively replaced by control computer software, a significant advantage in control capability and flexibility can be gained in future developmental testing, with practical implications for flight systems as well. Finally, it is concluded from computer simulations that TVCS stabilization using rate estimation promises a substantial performance improvement over the present design.

Fleischer, G. E.↗

Reliability Analysis for Unreliable FSM Computations

Finite State Machines (FSMs) are fundamental in both hardware design and software development. However, the reliability of FSM computations remains poorly understood. This becomes increasingly critical as the technology scaling becomes more aggressive. Existing reliability analyses are mainly designed for generic computations and are unaware of the special error tolerance characteristics in FSM computations. In this work, we introduce RelyFSM – a state-level reliability analysis framework for FSM computations. By qualitatively reasoning about the transition structures in FSM computations, RelyFSM can precisely capture their inherent error tolerance. Basically, RelyFSM takes an FSM specification as input and analyzes its reliability for a specific execution. At the core of the analysis is a set of mathematical models that accurately define the behaviors of unreliable FSM executions. Our evaluation with a spectrum of FSM benchmarks confirms the accuracy and efficiency of RelyFSM in analyzing the reliability of real-world FSM computations.

97 MATHEMATICS AND COMPUTING↗

Differentiable Neural Architecture, Mixed Precision and Accelerator Co-Search

Quantization, effective Neural Network architecture, and efficient accelerator hardware are three important design paradigms to maximize accuracy and efficiency. Mixed Precision Quantization is a process of assigning different precision to different Neural Network layers for optimized inference. Neural Architecture Search (NAS) is a process of automatically designing the neural network for a task and can also be extended to search for the precision of each weight and activation matrix. In this paper, we develop the following three methods: (i) Fast Differentiable Hardware-aware Mixed Precision Quantization Search method to find optimal precision, (ii) Joint Differentiable hardware-aware Architecture and Mixed Precision Quantization Co-search, (iii) Joint Accelerator, Architecture, and Precision triple co-search to find best possibilities in all the three worlds. We demonstrate the effectiveness of our proposed methods targeting Bitfusion accelerator by searching mixed precision models on MobilenetV2. We achieve better accuracy-latency trade-off models than the manually designed and previously proposed search methods.

97 MATHEMATICS AND COMPUTING↗

Testing a Neural Network Accelerator on a High-Altitude Balloon

The cognitive communications project has been working to re d machine learning approaches to support their deployment and sustained use in space environments. It has historically been difficult to implement such techniques on space platforms, however, due to the computational requirements they levy onto general-purpose avionics hardware. While technologies exist to accelerate the computation of aspects of neural networks, such platforms have not historically been deployed in space environments. Given that testing payloads in such environments can be both cost- and time-prohibitive, high-altitude balloons can be used as a way to approximate a space environment at a much lower cost, thus providing a cost-effective way in which to test newer approaches to hardware acceleration for artificial intelligence which may be deployed onto spacecraft more directly. This paper describes a successful test of a commercial off- the-shelf neural network accelerator on a high-altitude balloon. It begins by explaining our selection criteria when evaluating different commercial neural network acceleration techniques: primary considerations include size, weight, and power (SWaP) as well as ease of integration. Next, the paper describes the development and implementation of an experimental flight test platform: flight and ground components are discussed. Afterward, the paper discusses the experimental payload itself: this includes the experimental procedure as well as the specific image and method used for testing. Finally, the paper concludes with an evaluation of both the experimental device tested at altitude as well as the flight test framework itself, identifying how the existing platform can be used to continue tes g commercial off-the-shelf (COTS) solutions for acceleration.

Clark, Gilbert↗

DFSynthesizer: Dataflow-based Synthesis of Spiking Neural Networks to Neuromorphic Hardware

Spiking Neural Networks (SNNs) are an emerging computation model that uses event-driven activation and bio-inspired learning algorithms. SNN-based machine learning programs are typically executed on tile-based neuromorphic hardware platforms, where each tile consists of a computation unit called a crossbar, which maps neurons and synapses of the program. However, synthesizing such programs on an off-the-shelf neuromorphic hardware is challenging. This is because of the inherent resource and latency limitations of the hardware, which impact both model performance, e.g., accuracy, and hardware performance, e.g., throughput. We propose DFSynthesizer, an end-to-end framework for synthesizing SNN-based machine learning programs to neuromorphic hardware. The proposed framework works in four steps. First, it analyzes a machine learning program and generates SNN workload using representative data. Second, it partitions the SNN workload and generates clusters that fit on crossbars of the target neuromorphic hardware. Third, it exploits the rich semantics of the Synchronous Dataflow Graph (SDFG) to represent a clustered SNN program, allowing for performance analysis in terms of key hardware constraints such as number of crossbars, dimension of each crossbar, buffer space on tiles, and tile communication bandwidth. Finally, it uses a novel scheduling algorithm to execute clusters on crossbars of the hardware, guaranteeing hardware performance. We evaluate DFSynthesizer with 10 commonly used machine learning programs. Our results demonstrate that DFSynthesizer provides a much tighter performance guarantee compared to current mapping approaches.

Computer Science↗

EOS-AM precision pointing verification

The Earth Observing System (EOS) AM mission requires tight pointing knowledge to meet scientific objectives, in a spacecraft with low frequency flexible appendage modes. As the spacecraft controller reacts to various disturbance sources and as the inherent appendage modes are excited by this control action, verification of precision pointing knowledge becomes particularly challenging for the EOS-AM mission. As presently conceived, this verification includes a complementary set of multi-disciplinary analyses, hardware tests and real-time computer in the loop simulations, followed by collection and analysis of hardware test and flight data and supported by a comprehensive data base repository for validated program values.

Throckmorton, A.↗