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 1,621 records · Page 90

Implementing Molecular Dynamics on Hybrid High Performance Computers - Particle-Particle Particle-Mesh

The use of accelerators such as graphics processing units (GPUs) has become popular in scientific computing applications due to their low cost, impressive floating-point capabilities, high memory bandwidth, and low electrical power requirements. Hybrid high-performance computers, machines with nodes containing more than one type of floating-point processor (e.g. CPU and GPU), are now becoming more prevalent due to these advantages. In this paper, we present a continuation of previous work implementing algorithms for using accelerators into the LAMMPS molecular dynamics software for distributed memory parallel hybrid machines. In our previous work, we focused on acceleration for short-range models with an approach intended to harness the processing power of both the accelerator and (multi-core) CPUs. To augment the existing implementations, we present an efficient implementation of long-range electrostatic force calculation for molecular dynamics. Specifically, we present an implementation of the particle-particle particle-mesh method based on the work by Harvey and De Fabritiis. We present benchmark results on the Keeneland InfiniBand GPU cluster. We provide a performance comparison of the same kernels compiled with both CUDA and OpenCL. We discuss limitations to parallel efficiency and future directions for improving performance on hybrid or heterogeneous computers.

97 MATHEMATICS AND COMPUTING↗

Highly Efficient Compensation-Based Parallelism for Wavefront Loops on GPUs

Wavefront loops are widely used in many scientific applications, e.g., partial differential equation (PDE) solvers and sequence alignment tools. However, due to the data dependencies in wavefront loops, it is challenging to fully utilize the abundant compute units of GPUs and to reuse data through their memory hierarchy. Existing solutions can only optimize for these factors to a limited extent. For example, tiling-based methods optimize memory access but may result in load imbalance; while compensation-based methods, which change the original order of computation to expose more parallelism and then compensate for it, suffer from both global synchronization overhead and limited generality. In this paper, we first prove under which circumstances that breaking data dependencies and properly changing the sequence of computation operators in our compensation-based method does not affect the correctness of results. Based on this analysis, we design a highly efficient compensation-based parallelism on GPUs. Our method provides weighted scan-based GPU kernels to optimize the computation and combines with the tiling method to optimize memory access and synchronization. The performance results on the NVIDIA K80 and P100 GPU platforms demonstrate that our method can achieve significant improvements for four types of real-world application kernels over the state-of-the-art research.

Hou, Kaixi↗

Self-correcting quantum memory in a thermal environment

The ability to store information is of fundamental importance to any computer, be it classical or quantum. To identify systems for quantum memories, which rely, analogously to classical memories, on passive error protection (''self-correction''), is of greatest interest in quantum information science. While systems with topological ground states have been considered to be promising candidates, a large class of them was recently proven unstable against thermal fluctuations. Here, we propose two-dimensional (2D) spin models unaffected by this result. Specifically, we introduce repulsive long-range interactions in the toric code and establish a memory lifetime polynomially increasing with the system size. This remarkable stability is shown to originate directly from the repulsive long-range nature of the interactions. We study the time dynamics of the quantum memory in terms of diffusing anyons and support our analytical results with extensive numerical simulations. Our findings demonstrate that self-correcting quantum memories can exist in 2D at finite temperatures.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Tools Ensure Reliability of Critical Software

In November 2006, after attempting to make a routine maneuver, NASA's Mars Global Surveyor (MGS) reported unexpected errors. The onboard software switched to backup resources, and a 2-day lapse in communication took place between the spacecraft and Earth. When a signal was finally received, it indicated that MGS had entered safe mode, a state of restricted activity in which the computer awaits instructions from Earth. After more than 9 years of successful operation gathering data and snapping pictures of Mars to characterize the planet's land and weather communication between MGS and Earth suddenly stopped. Months later, a report from NASA's internal review board found the spacecraft's battery failed due to an unfortunate sequence of events. Updates to the spacecraft's software, which had taken place months earlier, were written to the wrong memory address in the spacecraft's computer. In short, the mission ended because of a software defect. Over the last decade, spacecraft have become increasingly reliant on software to carry out mission operations. In fact, the next mission to Mars, the Mars Science Laboratory, will rely on more software than all earlier missions to Mars combined. According to Gerard Holzmann, manager at the Laboratory for Reliable Software (LaRS) at NASA's Jet Propulsion Laboratory (JPL), even the fault protection systems on a spacecraft are mostly software-based. For reasons like these, well-functioning software is critical for NASA. In the same year as the failure of MGS, Holzmann presented a new approach to critical software development to help reduce risk and provide consistency. He proposed The Power of 10: Rules for Developing Safety-Critical Code, which is a small set of rules that can easily be remembered, clearly relate to risk, and allow compliance to be verified. The reaction at JPL was positive, and developers in the private sector embraced Holzmann's ideas.

Source record↗

Hutchinson Trace Estimation for high-dimensional and high-order Physics-Informed Neural Networks

Physics-Informed Neural Networks (PINNs) have proven effective in solving partial differential equations (PDEs), especially when some data are available by seamlessly blending data and physics. However, extending PINNs to high-dimensional and even high-order PDEs encounters significant challenges due to the computational cost associated with automatic differentiation in the residual loss function calculation. Herein, we address the limitations of PINNs in handling high-dimensional and high-order PDEs by introducing the Hutchinson Trace Estimation (HTE) method. Starting with the second-order high-dimensional PDEs, which are ubiquitous in scientific computing, HTE is applied to transform the calculation of the entire Hessian matrix into a Hessian vector product (HVP). This approach not only alleviates the computational bottleneck via Taylor-mode automatic differentiation but also significantly reduces memory consumption from the Hessian matrix to an HVP’s scalar output. We further showcase HTE’s convergence to the original PINN loss and its unbiased behavior under specific conditions. Comparisons with the Stochastic Dimension Gradient Descent (SDGD) highlight the distinct advantages of HTE, particularly in scenarios with significant variability and variance among dimensions. We further extend the application of HTE to higher-order and higher-dimensional PDEs, specifically addressing the biharmonic equation. By employing tensor-vector products (TVP), HTE efficiently computes the colossal tensor associated with the fourth-order high-dimensional biharmonic equation, saving memory and enabling rapid computation. The effectiveness of HTE is illustrated through experimental setups, demonstrating comparable convergence rates with SDGD under memory and speed constraints. Additionally, HTE proves valuable in accelerating the Gradient-Enhanced PINN (gPINN) version as well as the Biharmonic equation. Overall, HTE opens up a new capability in scientific machine learning for tackling high-order and high-dimensional PDEs.

Curse of dimensionality↗

Apollo guidance, navigation and control: Guidance system operations plans for manned LM earth orbital and lunar missions using Program COLOSSUS 3. Section 7: Erasable memory programs

Erasable-memory programs (EMPs) designed for the guidance computers used in the command (CMC) and lunar modules (LGC) are described. CMC programs are designated COLOSSUS 3, and the associated EMPs are identified by a three-digit number beginning with 5. LGC programs are designated LUMINARY 1E, and the associated EMPs are identified, with one exception, by a three-digit number beginning with 1. The exception is EMP 99. The EMPs vary in complexity from a simple flagbit setting to a long and intricate logical structure. They all, however, cause the computer to behave in a way not intended in the original design of the programs; they accomplish this off-nominal behavior by some alteration of erasable memory to interface with existing fixed-memory programs to effect a desired result.

Hamilton, M. H.↗

Distributed seams for gigapixel panoramas

Gigapixel panoramas are an increasingly popular digital image application. They are often created as a mosaic of many smaller images. The mosaic acquisition can take many hours causing the individual images to differ in exposure and lighting conditions. A blending operation is often necessary to give the appearance of a seamless image. The blending quality depends on the magnitude of discontinuity along the image boundaries. Often, new boundaries, or seams, are first computed that minimize this transition. Current techniques based on multi-labeling Graph Cuts are too slow and memory intensive for gigapixel sized panoramas. In this paper, we present a parallel, out-of-core seam computing technique that is fast, has small memory footprint, and is capable of running efficiently on different types of parallel systems. Its maximum memory usage is configurable, in the form of a cache, which can improve performance by reducing redundant disk I/O and computations. It shows near-perfect scaling on symmetric multiprocessing systems and good scaling on clusters and distributed shared memory systems. Lastly, our technique improves the time required to compute seams for gigapixel imagery from many hours (or even days) to just a few minutes, while still producing boundaries with energy that is on-par with Graph Cuts.

97 MATHEMATICS AND COMPUTING↗

Cycle accurate and cycle reproducible memory for an FPGA based hardware accelerator

A method, system and computer program product are disclosed for using a Field Programmable Gate Array (FPGA) to simulate operations of a device under test (DUT). The DUT includes a device memory having a number of input ports, and the FPGA is associated with a target memory having a second number of input ports, the second number being less than the first number. In one embodiment, a given set of inputs is applied to the device memory at a frequency Fd and in a defined cycle of time, and the given set of inputs is applied to the target memory at a frequency Ft. Ft is greater than Fd and cycle accuracy is maintained between the device memory and the target memory. In an embodiment, a cycle accurate model of the DUT memory is created by separating the DUT memory interface protocol from the target memory storage array.

42 ENGINEERING↗

Active Flash: Out-of-core Data Analytics on Flash Storage

Next generation science will increasingly come to rely on the ability to perform efficient, on-the-fly analytics of data generated by high-performance computing (HPC) simulations, modeling complex physical phenomena. Scientific computing workflows are stymied by the traditional chaining of simulation and data analysis, creating multiple rounds of redundant reads and writes to the storage system, which grows in cost with the ever-increasing gap between compute and storage speeds in HPC clusters. Recent HPC acquisitions have introduced compute node-local flash storage as a means to alleviate this I/O bottleneck. We propose a novel approach, Active Flash, to expedite data analysis pipelines by migrating to the location of the data, the flash device itself. We argue that Active Flash has the potential to enable true out-of-core data analytics by freeing up both the compute core and the associated main memory. By performing analysis locally, dependence on limited bandwidth to a central storage system is reduced, while allowing this analysis to proceed in parallel with the main application. In addition, offloading work from the host to the more power-efficient controller reduces peak system power usage, which is already in the megawatt range and poses a major barrier to HPC system scalability. We propose an architecture for Active Flash, explore energy and performance trade-offs in moving computation from host to storage, demonstrate the ability of appropriate embedded controllers to perform data analysis and reduction tasks at speeds sufficient for this application, and present a simulation study of Active Flash scheduling policies. These results show the viability of the Active Flash model, and its capability to potentially have a transformative impact on scientific data analysis.

99 GENERAL AND MISCELLANEOUS↗

Parameter optimization for transitions between memory states in small arrays of Josephson junctions

Coupled arrays of Josephson junctions possess multiple stable zero voltage states. Such states can store information and consequently can be utilized for cryogenic memory applications. Basic memory operations can be implemented by sending a pulse to one of the junctions and studying transitions between the states. In order to be suitable for memory operations, such transitions between the states have to be fast and energy efficient. Here in this article we employed simulated annealing, a stochastic optimization algorithm, to study parameter optimization of array parameters which minimizes times and energies of transitions between specifically chosen states that can be utilized for memory operations (Read, Write, and Reset). Simulation results show that such transitions occur with access times on the order of 10–100 ps and access energies on the order of 10 -19 –5×10 -18 J. Numerical simulations are validated with approximate analytical results.

97 MATHEMATICS AND COMPUTING↗

New modelling capabilities in IDT

This work concerns the enhanced modelling capabilities of the discrete ordinates transport solver IDT. The novelties introduced allow for modelling unstructured geometries composed by a collection of X/Y segments and circles, and the use of reciprocity and conservation relations reduce the memory imprint as well as the computational cost of the method. IDT decomposes geometries in modular Cartesian patterns, which are the so-called Heterogeneous Cartesian Cells (HCCs), containing a chunk of the original unstructured geometries. Each HCC can be then discretized by superimposing a XY grid to refine locally the HCC. Unlike the most popular MOC, IDT performs the spatial sweeping by directional collision probabilities instead of trajectories. The sources and interface angular fluxes are expanded up to linear order. The accuracy of ray-tracing, the memory imprints together with the novel mesh refinement capabilities have been verified. A first set of preliminary results on PWR lattice problems will be presented in this paper.

22 GENERAL STUDIES OF NUCLEAR REACTORS↗

A fast band–Krylov eigensolver for macromolecular functional motion simulation on multicore architectures and graphics processors

We introduce a new iterative Krylov subspace-based eigensolver for the simulation of macromolecular motions on desktop multithreaded platforms equipped with multicore processors and, possibly, a graphics accelerator (GPU). The method consists of two stages, with the original problem first reduced into a simpler band-structured form by means of a high-performance compute-intensive procedure. This is followed by a memory-intensive but low-cost Krylov iteration, which is off-loaded to be computed on the GPU by means of an efficient data-parallel kernel. The experimental results reveal the performance of the new eigensolver. Concretely, when applied to the simulation of macromolecules with a few thousands degrees of freedom and the number of eigenpairs to be computed is small to moderate, the new solver outperforms other methods implemented as part of high-performance numerical linear algebra packages for multithreaded architectures.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Simulations of transition and turbulence on the Navier-Stokes computer

The Navier-Stokes Computer (NSC) consists of multiple local memory parallel processors interconnected in a hypercube network. Efficient implementation of algorithms on the NSC thus requires the effective utilization of both the coarse and fine grain paralelism inherent in the architectural design. The basic approach to implementing an algorithm on the NSC is presented herein. The particular finite-difference algorithm considered was developed for performing transition and turbulence simulations by direct solution of the time-dependent incompressible Navier-Stokes equations. The suitability of this algorithm for performing simulations of the isotropic turbulence problem is verified from computations performed on a Cray 2. Projected timing results for the algorithm on the NSC itself are presented for both the isotropic turbulence and laminar turbulent transition problems.

Krist, S. E.↗

DMA engine for repeating communication patterns

A parallel computer system is constructed as a network of interconnected compute nodes to operate a global message-passing application for performing communications across the network. Each of the compute nodes includes one or more individual processors with memories which run local instances of the global message-passing application operating at each compute node to carry out local processing operations independent of processing operations carried out at other compute nodes. Each compute node also includes a DMA engine constructed to interact with the application via Injection FIFO Metadata describing multiple Injection FIFOs where each Injection FIFO may containing an arbitrary number of message descriptors in order to process messages with a fixed processing overhead irrespective of the number of message descriptors included in the Injection FIFO.

Chen, Dong↗

Parallel Processing of Geospatial Time-Series Data

One way of quantifying how much sunlight the earth receives, at a particular location and time, is to compute a Cloud Index from satellite images. This calculation involves processing large files of data that are collected in a way that makes time-series analysis extremely time consuming on a single computer. There are several ways to render this memory-intensive problem more feasible usingparallel programming. Our implementation uses the Message Passing Interface (MPI) protocol and is an excellent balance between implementation complexity and execution time. We believe that this will also scale well as more complex algorithms for computing the Cloud Index emerge. This work extends to any domain where time-series analysis is needed on large data sets that are collectedas a function of time.

97 MATHEMATICS AND COMPUTING↗

Improving Memory Error Handling Using Linux

As supercomputers continue to get faster and more powerful in the future, they will also have more nodes. If nothing is done, then the amount of memory in supercomputer clusters will soon grow large enough that memory failures will be unmanageable to deal with by manually replacing memory DIMMs. "Improving Memory Error Handling Using Linux" is a process oriented method to solve this problem by using the Linux kernel to disable (offline) faulty memory pages containing bad addresses, preventing them from being used again by a process. The process of offlining memory pages simplifies error handling and results in reducing both hardware and manpower costs required to run Los Alamos National Laboratory (LANL) clusters. This process will be necessary for the future of supercomputing to allow the development of exascale computers. It will not be feasible without memory error handling to manually replace the number of DIMMs that will fail daily on a machine consisting of 32-128 petabytes of memory. Testing reveals the process of offlining memory pages works and is relatively simple to use. As more and more testing is conducted, the entire process will be automated within the high-performance computing (HPC) monitoring software, Zenoss, at LANL.

97 MATHEMATICS AND COMPUTING↗

Realization of Universal Ion-Trap Quantum Computation with Decoherence-Free Qubits

Any residual coupling of a quantum computer to the environment results in computational errors. Encoding quantum information in a so-called decoherence-free subspace provides means to avoid these errors. Despite tremendous progress in employing this technique to extend memory storage times by orders of magnitude, computation within such subspaces has been scarce. Here, we demonstrate the realization of a universal set of quantum gates acting on decoherence-free ion qubits. We combine these gates to realize the first controlled-NOT gate towards a decoherence-free, scalable quantum computer.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Electrochemical Random-Access Memory: Progress, Perspectives, and Opportunities

Non-von Neumann computing using neuromorphic systems based on analogue synaptic and neuronal elements has emerged as a potential solution to tackle the growing need for more efficient data processing, but progress toward practical systems has been stymied due to a lack of materials and devices with the appropriate attributes. Recently, solid state electrochemical ion-insertion, also known as electrochemical random access memory (ECRAM) has emerged as a promising approach to realize the needed device characteristics. ECRAM is a three terminal device that operates by tuning electronic conductance in functional materials through solid-state electrochemical redox reactions. This mechanism can be considered as a gate-controlled bulk modulation of dopants and/or phases in the channel. Early work demonstrating that ECRAM can achieve nearly ideal analogue synaptic characteristics has sparked tremendous interest in this approach. More recently, the realization that electrochemical ion insertion can be used to tune the electronic properties of many types of materials including transition metal oxides, layered two-dimensional materials, organic and coordination polymers, and that the changes in conductance can span orders of magnitude has further attracted interest in ECRAM as the basis for analogue synaptic elements for inference accelerators as well as for dynamical devices that can emulate a wide range of neuronal characteristics for implementation in analogue spiking neural networks. At its core, ECRAM shares many fundamental aspects with rechargeable batteries, where ion insertion materials are used extensively for their ability to reversibly store charge and energy. Computing applications, however, present drastically different requirements: systems will require many millions of devices, scaled down to tens of nanometers, all while achieving reliable electronic-state tuning at scaled-up rates and endurances, and with minimal energy dissipation and noise. Further, in this review, we discuss the history, basic concepts, recent progress, as well as the challenges and opportunities for different types of ECRAM, broadly grouped by their primary mobile ionic charge carrier, including Li, protons, and oxygen vacancies.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗