Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “compute in 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 1,765 records · Page 98

System applications of the fault tolerant memory

Conventional memory technologies currently employed in aerospace applications contribute at least fifty percent to system unreliability (where the system includes CPU, I/O and memory). A fault tolerant memory performs both error correction and memory replacement at the bit plane level. To determine the effects of system design of using a fault tolerant memory in space applications, analysis was performed to determine tradeable hardware configurations that meet the reliability goals of each program. The candidate configurations, which included redundant elements of the computer system with both conventional and fault tolerant memories, were then traded in terms of selection criteria of cost, weight, volume, and power. These trade studies demonstrated that a fault tolerant memory provided significant advantages in terms of cost, weight, and volume. The memory selected for this analysis was a recently developed five fault tolerant memory.

Murphy, L. J.↗

Samba: A Detailed Memory Management Unit (MMU) for the SST Simulation Framework

The Memory Management Unit (MMU) is one of the most important parts of any modern computing system. It can be thought of as the hardware support for virtual memory, which enforces access permissions, and manages the translation of processes' virtual addresses into real memory physical addresses. The operating system (OS) allocates the physical memory in a small granularity frames that can be accessed by processes with appropriate permissions. Each process has the illusion of having an entire memory space, however, the actual addresses issued by the process, (`virtual addresses'), are translated into the physical frame address allocated by the OS through virtual memory.

97 MATHEMATICS AND COMPUTING↗

Enhancing scalability of a matrix-free eigensolver for studying many-body localization

We propose several techniques to enhance the parallel scalability of a matrix-free eigensolver designed for studying many-body localization (MBL) of quantum spin chain models with nearest-neighbor interactions and on-site disorder. This type of problem is computationally challenging because the dimension of the associated Hamiltonian matrix grows exponentially with respect to the number of spins L, and we need to average over different realizations of the random disorder to obtain relevant statistical behavior. For each disorder realization, we need to compute eigenvalues from different regions of the spectrum and their corresponding eigenvectors. In previous work, the interior eigenstates for a single eigenvalue problem are computed via the shift-and-invert Lanczos algorithm. Due to the extremely high memory footprint of the LU factorizations, this technique is not well suited for large L’s. For example, we need thousands of compute nodes on modern high performance computing infrastructures to go beyond L = 24. The matrix-free approach does not suffer from this memory bottleneck, however, its scalability is limited by a computation and communication load imbalance. To reduce this imbalance and to significantly enhance the scalability of the matrix-free eigensolver, we reorder the matrix and leverage the consistent space runtime, CSPACER. We also show its efficiency in managing irregular communication patterns at scale compared to optimized MPI non-blocking two-sided and one-sided RMA implementation variants. This effort enables us to study MBL for spin chains with a larger number of spins. The efficiency and effectiveness of the proposed algorithm is demonstrated by computing eigenstates on a massively parallel many-core high performance computer.

METIS↗

RighTime: A real time clock correcting program for MS-DOS-based computer systems

A computer program is described which effectively eliminates the misgivings of the DOS system clock in PC/AT-class computers. RighTime is a small, sophisticated memory-resident program that automatically corrects both the DOS system clock and the hardware 'CMOS' real time clock (RTC) in real time. RighTime learns what corrections are required without operator interaction beyond the occasional accurate time set. Both warm (power on) and cool (power off) errors are corrected, usually yielding better than one part per million accuracy in the typical desktop computer with no additional hardware, and RighTime increases the system clock resolution from approximately 0.0549 second to 0.01 second. Program tools are also available which allow visualization of RighTime's actions, verification of its performance, display of its history log, and which provide data for graphing of the system clock behavior. The program has found application in a wide variety of industries, including astronomy, satellite tracking, communications, broadcasting, transportation, public utilities, manufacturing, medicine, and the military.

Becker, G. Thomas↗

Challenges Facing Design and Analysis Tools

The design and analysis of future aerospace systems will strongly rely on advanced engineering analysis tools used in combination with risk mitigation procedures. The implications of such a trend place increased demands on these tools to assess off-nominal conditions, residual strength, damage propagation, and extreme loading conditions in order to understand and quantify these effects as they affect mission success. Advances in computer hardware such as CPU processing speed, memory, secondary storage, and visualization provide significant resources for the engineer to exploit in engineering design. The challenges facing design and analysis tools fall into three primary areas. The first area involves mechanics needs such as constitutive modeling, contact and penetration simulation, crack growth prediction, damage initiation and progression prediction, transient dynamics and deployment simulations, and solution algorithms. The second area involves computational needs such as fast, robust solvers, adaptivity for model and solution strategies, control processes for concurrent, distributed computing for uncertainty assessments, and immersive technology. Traditional finite element codes still require fast direct solvers which when coupled to current CPU power enables new insight as a result of high-fidelity modeling. The third area involves decision making by the analyst. This area involves the integration and interrogation of vast amounts of information - some global in character while local details are critical and often drive the design. The proposed presentation will describe and illustrate these areas using composite structures, energy-absorbing structures, and inflatable space structures. While certain engineering approximations within the finite element model may be adequate for global response prediction, they generally are inadequate in a design setting or when local response prediction is critical. Pitfalls to be avoided and trends for emerging analysis tools will be described.

Knight, Norman F., Jr.↗

Post Moore's Law Report

Moore's law is driving an information revolution, worldwide economic growth, and is a tool for national security. This report explains how dire proclamations that "Moore's law is ending" are due to a natural redefinition of the phrase, but computing remains positioned to both drive economic growth and support national security. The computer industry used to be led by the semiconductor companies that made ever faster microprocessors with larger memories. However, control is shifting to new ways of designing computers, notably based on 3D chips and new analog and digital architectures. While artificial intelligence and quantum computing research have become mainstream pursuits, these latter two areas seem destined split off from Moore's law rather than become a part of it. We include a discussion of recent developments and opportunities in optical communications and computing.

97 MATHEMATICS AND COMPUTING↗

Basic memory module

Construction and electrical characterization of the 4096 x 2-bit Basic Memory Module (BMM) are reported for the Space Ultrareliable Modular Computer (SUMC) program. The module uses four 2K x 1-bit N-channel FET, random access memory chips, called array chips, and two sense amplifier chips, mounted and interconnected on a ceramic substrate. Four 5% tolerance power supplies are required. At the Module, the address, chip select, and array select lines require a 0-8.5 V MOS signal level. The data output, read-strobe, and write-enable lines operate at TTl levels. Although the module is organized as 4096 x 2 bits, it can be used in a 8196 x 1-bit application with appropriate external connections. A 4096 x 1-bit organization can be obtained by depopulating chips.

Tietze, F. C.↗

Addressing Inter-set Write-Variation for Improving Lifetime of Non-Volatile Caches

We propose a technique which minimizes inter-set write variation in NVM caches for improving its lifetime. Our technique uses cache coloring scheme to add a software-controlled mapping layer between groups of physical pages (called memory regions) and cache sets. Periodically, the number of writes to different colors of the cache is computed and based on this result, the mapping of a few colors is changed to channel the write traffic to least utilized cache colors. This change helps to achieve wear-leveling.

cache↗

tomoCAM : fast model-based iterative reconstruction via GPU acceleration and non-uniform fast Fourier transforms

X-ray-based computed tomography is a well established technique for determining the three-dimensional structure of an object from its two-dimensional projections. In the past few decades, there have been significant advancements in the brightness and detector technology of tomography instruments at synchrotron sources. These advancements have led to the emergence of new observations and discoveries, with improved capabilities such as faster frame rates, larger fields of view, higher resolution and higher dimensionality. These advancements have enabled the material science community to expand the scope of tomographic measurements towards increasingly in situ and in operando measurements. In these new experiments, samples can be rapidly evolving, have complex geometries and restrictions on the field of view, limiting the number of projections that can be collected. In such cases, standard filtered back-projection often results in poor quality reconstructions. Iterative reconstruction algorithms, such as model-based iterative reconstructions (MBIR), have demonstrated considerable success in producing high-quality reconstructions under such restrictions, but typically require high-performance computing resources with hundreds of compute nodes to solve the problem in a reasonable time. Here, tomoCAM , is introduced, a new GPU-accelerated implementation of model-based iterative reconstruction that leverages non-uniform fast Fourier transforms to efficiently compute Radon and back-projection operators and asynchronous memory transfers to maximize the throughput to the GPU memory. The resulting code is significantly faster than traditional MBIR codes and delivers the reconstructive improvement offered by MBIR with affordable computing time and resources. tomoCAM has a Python front-end, allowing access from Jupyter -based frameworks, providing straightforward integration into existing workflows at synchrotron facilities.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

Optical Addressing And Clocking Of RAM's

Proposed random-access-memory (RAM) addressing system, in which memory linked optically to read/write logic circuits, greatly increases computer operating speed. System - comprises addressing circuits including numerous lasers as signal sources, numerous optical gates including optical detectors associated with memory cells, and holographic element to direct light signals to desired memory-cell locations - applied to high-capacity digital systems, supercomputers, and complex microcircuits.

Johnston, Alan R.↗

Digital Instrumentation Analysis and Navigation System (DIANS) for system identification

Hardware, software, and performance features of the Digital Instrumentation and Navigation System (DIANS) designed for NASA research to collect flight data as a strap-down system are detailed. The support software for the system has a cross compiler, a linkage editor, and cross assembler, extended communication capabilities, postflight processing applications, and compilers for PASCAL, FORTRAN, CBASIC and MT. The DIANS microcomputer has a 1 Mbyte RAM module, a fast floating point processor board, a 68000 monoboard computer with 64 RAM, a 128 Kbyte bubble memory card, and a navigation radio. The system also carries a battery for full system operation for over an hour. The support software is also stored on a host mainframe computer, which has a CP/M operating system. Pitch, roll, and heading data are gathered from the on-board system, and communication is possible between the airborne and ground-based computer.

Smyth, R. K.↗

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↗

NVMalloc: Exposing an Aggregate SSD Store as a Memory Partition in Extreme-Scale Machines

DRAM is a precious resource in extreme-scale machines and is increasingly becoming scarce, mainly due to the growing number of cores per node. On future multi-petaflop and exaflop machines, the memory pressure is likely to be so severe that we need to rethink our memory usage models. Fortunately, the advent of non-volatile memory (NVM) offers a unique opportunity in this space. Current NVM offerings possess several desirable properties, such as low cost and power efficiency, but suffer from high latency and lifetime issues. We need rich techniques to be able to use them alongside DRAM. In this talk, we present a novel approach for exploiting NVM as a secondary memory partition so that applications can explicitly allocate and manipulate memory regions therein. More specifically, we propose an NVMalloc library with a suite of services that enables applications to access a distributed NVM storage system. We have devised ways within NVMalloc so that the storage system, built from compute node-local NVM devices, can be accessed in a byte-addressable fashion using the memory mapped I/O interface. Our approach has the potential to re-energize out-of-core computations on largescale machines by having applications allocate certain variables through NVMalloc, thereby increasing the overall memory capacity available. Our evaluation on a 128-core cluster shows that NVMalloc enables applications to compute problem sizes larger than the physical memory in a cost-effective manner. It can bring more performance/efficiency gain with increased computation time between NVM memory accesses or increased data access locality. In addition, our results suggest that while NVMalloc enables transparent access to NVM-resident variables, the explicit control it provides is crucial to optimize application performance.

99 GENERAL AND MISCELLANEOUS↗

Systems and methods for rapid processing and storage of data

Systems and methods of building massively parallel computing systems using low power computing complexes in accordance with embodiments of the invention are disclosed. A massively parallel computing system in accordance with one embodiment of the invention includes at least one Solid State Blade configured to communicate via a high performance network fabric. In addition, each Solid State Blade includes a processor configured to communicate with a plurality of low power computing complexes interconnected by a router, and each low power computing complex includes at least one general processing core, an accelerator, an I/O interface, and cache memory and is configured to communicate with non-volatile solid state memory.

97 MATHEMATICS AND COMPUTING↗

Accelerating k-NN Algorithm with Hybrid MPI and OpenSHMEM

Machine Learning algorithms are benefiting from the continuous improvement of programming models, including MPI, MapReduce and PGAS. k-Nearest Neighbors (k-NN) algorithm is a widely used machine learning algorithm, applied to supervised learning tasks such as classification. Several parallel implementations of k-NN have been proposed in the literature and practice. However, on high-performance computing systems with high-speed interconnects, it is important to further accelerate existing designs of the k-NN algorithm through taking advantage of scalable programming models. To improve the performance of k-NN on large-scale environment with InfiniBand network, this paper proposes several alternative hybrid MPI+OpenSHMEM designs and performs a systemic evaluation and analysis on typical workloads. The hybrid designs leverage the one-sided memory access to better overlap communication with computation than the existing pure MPI design, and propose better schemes for efficient buffer management. The implementation based on k-NN program from MaTEx with MVAPICH2-X (Unified MPI+PGAS Communication Runtime over InfiniBand) shows up to 9.0% time reduction for training KDD Cup 2010 workload over 512 cores, and 27.6% time reduction for small workload with balanced communication and computation. Experiments of running with varied number of cores show that our design can maintain good scalability.

Lin, Jian↗

NVMalloc: Exposing an Aggregate SSD Store as a Memory Partition in Extreme-Scale Machines

DRAM is a precious resource in extreme-scale machines and is increasingly becoming scarce, mainly due to the growing number of cores per node. On future multi-petaflop and exaflop machines, the memory pressure is likely to be so severe that we need to rethink our memory usage models. Fortunately, the advent of non-volatile memory (NVM) offers a unique opportunity in this space. Current NVM offerings possess several desirable properties, such as low cost and power efficiency, but suffer from high latency and lifetime issues. We need rich techniques to be able to use them alongside DRAM. In this paper, we propose a novel approach for exploiting NVM as a secondary memory partition so that applications can explicitly allocate and manipulate memory regions therein. More specifically, we propose an NVMalloc library with a suite of services that enables applications to access a distributed NVM storage system. We have devised ways within NVMalloc so that the storage system, built from compute node-local NVM devices, can be accessed in a byte-addressable fashion using the memory mapped I/O interface. Our approach has the potential to re-energize out-of-core computations on large-scale machines by having applications allocate certain variables through NVMalloc, thereby increasing the overall memory capacity available. Our evaluation on a 128-core cluster shows that NVMalloc enables applications to compute problem sizes larger than the physical memory in a cost-effective manner. It can bring more performance/efficiency gain with increased computation time between NVM memory accesses or increased data access locality. In addition, our results suggest that while NVMalloc enables transparent access to NVM-resident variables, the explicit control it provides is crucial to optimize application performance.

Computer Science↗

Efficient Training of Deep Neural Operator Networks via Randomized Sampling

Neural operators (NOs) employ deep neural networks to learn the mappings between infinitedimensional function spaces. Deep operator network (DeepONet), a popular NO architecture, has demonstrated success in the real-time prediction of complex dynamics across various scientific and engineering applications. In this work, we introduce a random sampling technique to be adopted during the training of DeepONet, aimed at improving the generalization ability of the model, while significantly reducing the computational time. The proposed approach targets the trunk network of the DeepONet model that outputs the basis functions corresponding to the spatiotemporal locations of the bounded domain on which the physical system is defined. While constructing the loss function, DeepONet training traditionally considers a uniform grid of spatiotemporal points at which all the output functions are evaluated for each iteration. This approach leads to a larger batch size, resulting in poor generalization and increased memory demands, due to the limitations of the stochastic gradient descent (SGD) optimizer. The proposed random sampling over the inputs of the trunk net mitigates these challenges, improving generalization and reducing the memory requirements during training, resulting in significant computational gains. We validate our hypothesis through three benchmark examples, demonstrating substantial reductions in training time while achieving comparable or lower overall test errors relative to the traditional training approach. Our results indicate that incorporating randomization in the trunk network inputs during training enhances the efficiency and robustness of DeepONet, offering a promising avenue for improving the framework’s performance in modeling complex physical systems.

Karumuri, Sharmila [Department of Civil & Systems ↗

Impacts of Hybrid Parallelism and Vectorization on the Performance of Newton-Krylov Methods in Computational Aerodynamics

Finding the numerical solution of moderate and high-fidelity aerodynamics problems on modern computer architectures involves, 1) decomposing the domain into smaller regions of nearly equal size, and 2) allocating computational resources for calculations on each domain and communication between domains. Modern computer clusters are composed from hierarchies of processing, memory, and communication resources with varying capabilities and latencies.This paper focuses on the combination of domain decomposition provided by ParMETIS [1]and Newton-Krylov Methods [2–5] for the solution of Computational Aerodynamics problems of interest to NASA. Herein, trade-offs encountered when mapping aerodynamics problems to modern computer architectures are explored through examples and discussions of trade-offs in parallelism from MPI [6], Open MP [7], and vectorization as partition sizes and computational resources are varied. An example of the impact that domain decomposition and MPI+OpenMPresource allocation can have on an adjoint calculation is presented in this abstract. The full paper will include more detailed examples, discussions of difficulties and potential methods to overcome them, and topics identified for future study.

Computational Aerodynamics, Hybrid Parallelism, Ve↗