Search NASA⌕ Search

SEARCH · Search NASA

Results for “parallel, scaling”

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 1,009 records · Page 56
All sourcesNASA NTRSDOE OSTI
NASA NTRS1991

The NAS parallel benchmarks

A new set of benchmarks has been developed for the performance evaluation of highly parallel supercomputers in the framework of the NASA Ames Numerical Aerodynamic Simulation (NAS) Program. These consist of five 'parallel kernel' benchmarks and three 'simulated application' benchmarks. Together they mimic the computation and data movement characteristics of large-scale computational fluid dynamics applications. The principal distinguishing feature of these benchmarks is their 'pencil and paper' specification-all details of these benchmarks are specified only algorithmically. In this way many of the difficulties associated with conventional benchmarking approaches on highly parallel systems are avoided.

Bailey, D. H.↗
DOE OSTI2020

Reducing Communication in Graph Neural Network Training

Graph Neural Networks (GNNs) are powerful and flexible neural networks that use the naturally sparse connectivity information of the data. GNNs represent this connectivity as sparse matrices, which have lower arithmetic intensity and thus higher communication costs compared to dense matrices, making GNNs harder to scale to high concurrencies than convolutional or fully-connected neural networks. Here, we introduce a family of parallel algorithms for training GNNs and show that they can asymptotically reduce communication compared to previous parallel GNN training methods. We implement these algorithms, which are based on 1D, 1. 5D, 2D, and 3D sparse-dense matrix multiplication, using torch.distributed on GPU-equipped clusters. Our algorithms optimize communication across the full GNN training pipeline. We train GNNs on over a hundred GPUs on multiple datasets, including a protein network with over a billion edges.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2020

The organic redox transistor for neuromorphic computing

Inspired by the in-memory computing architectures of biological systems, neuromorphic computing using crossbar arrays of artificial synapses based on non-volatile memory (NVM) devices with variable conductances has emerged as a new paradigm to enable massively parallel and ultra-low power computing hardware for data centric applications. Although inference has been demonstrated successfully using crossbars based on a variety of NMV technologies, efficient learning and scaling to large arrays (>10 6 elements) remains a challenge due to the synaptic elements' non-ideal electrical characteristics which degrades ANN accuracy. A further challenge is that in the conductive state memristors draw large currents >μA resulting in significant voltage drops in the interconnect wires and increased probability of failure in scaled arrays. We suggest the organic polymer redox transistor (RT) is an alternate approach that could solve many of these challenges, enabling both inference and parallel outer product updates, as recently demonstrated by Fuller et al. An RT consists of redox-active channel and gate electrodes in contact with a liquid or solid electrolyte. lon insertion through the electrolyte controls the channel electronic conductivity, while electron transfer through an external circuit maintains overall charge neutrality. Unlike a rechargeable battery, in the RT the voltage built-up across the electrolyte is kept to a minimum (typically <100 mV) by using the same material for the gate and channel. Elimination of the voltage offset simplifies integration of the RT into programmable arrays by enabling the use of various selectors. RTs based on inorganic and organic materials have been recently demonstrated with conductance tuning occurring at potentials of just a few mV and hundreds to thousands of linearly and symmetrically programmable conductance states, enabling near ideal accuracy in neural network simulations. Introduced in the 1980's, redox transistors with metallic gate electrodes and organic channel materials, also known as organic electrochemical transistors (OECTs), have been explored for a variety of applications such as chem- and bio-sensing, neural interfaces, and low cost printed circuits. A typical channel material for OECTs is the conducting polymer poly(3,4-ethylenedioxythiophene) doped with poly(styrene sulfonate) (PEDOT:PSS). PEDOT is a p-type semiconducting polymer with mobile positively charged polarons that hop chain-to-chain.

Talin, Albert Alec↗
DOE OSTI2017

A performance analysis of ensemble averaging for high fidelity turbulence simulations at the strong scaling limit

We analyze the potential performance benefits of estimating expected quantities in large eddy simulations of turbulent flows using true ensembles rather than ergodic time averaging. Multiple realizations of the same flow are simulated in parallel, using slightly perturbed initial conditions to create unique instantaneous evolutions of the flow field. Each realization is then used to calculate statistical quantities. Provided each instance is sufficiently de-correlated, this approach potentially allows considerable reduction in the time to solution beyond the strong scaling limit for a given accuracy. This study focuses on the theory and implementation of the methodology in Nek5000, a massively parallel open-source spectral element code.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
NASA NTRS1993

Resource Management for Distributed Parallel Systems

Multiprocessor systems should exist in the the larger context of distributed systems, allowing multiprocessor resources to be shared by those that need them. Unfortunately, typical multiprocessor resource management techniques do not scale to large networks. The Prospero Resource Manager (PRM) is a scalable resource allocation system that supports the allocation of processing resources in large networks and multiprocessor systems. To manage resources in such distributed parallel systems, PRM employs three types of managers: system managers, job managers, and node managers. There exist multiple independent instances of each type of manager, reducing bottlenecks. The complexity of each manager is further reduced because each is designed to utilize information at an appropriate level of abstraction.

Neuman, B. Clifford↗
DOE OSTI2025

Low-Energy Backgrounds in Solid-State Phonon and Charge Detectors

Solid-state phonon and charge detectors probe the scattering of weakly interacting particles, such as dark matter and neutrinos, through their low recoil thresholds. Recent advancements have pushed sensitivity to eV-scale energy depositions, uncovering previously-unseen low-energy excess backgrounds. While some arise from known processes such as thermal radiation, luminescence, and stress, others remain unexplained. This review examines these backgrounds, their possible origins, and parallels to low-energy effects in solids. Their understanding is essential for interpreting particle interactions at and below the eV-scale.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗
NASA NTRS1989

Adaptive implicit-explicit and parallel element-by-element iteration schemes

Adaptive implicit-explicit (AIE) and grouped element-by-element (GEBE) iteration schemes are presented for the finite element solution of large-scale problems in computational mechanics and physics. The AIE approach is based on the dynamic arrangement of the elements into differently treated groups. The GEBE procedure, which is a way of rewriting the EBE formulation to make its parallel processing potential and implementation more clear, is based on the static arrangement of the elements into groups with no inter-element coupling within each group. Various numerical tests performed demonstrate the savings in the CPU time and memory.

Tezduyar, T. E.↗
NASA NTRS1988

Gray scale operation of a multichannel optical convolver using the Semetex magnetooptic spatial light modulator

A new multichannel optical correlator/convolver architecture which uses an acoustooptic light modulator for the input channel and a Semetex magnetooptic spatial light modulator (MOSLM) for the set of parallel reference channels is presented. Details of the anamorphic optical system are discussed. Experimental results illustrate the use of the system as a convolver for performing digital multiplication by analog convolution (DMAC). A limited gray scale capability for data stored by the MOSLM is demonstrated by implementing this DMAC algorithm with trinary logic. Use of the MOSLM allows the number of parallel channels for the convolver to be increased significantly compared with previously reported techniques while retaining the capability for updating both channels at high speeds.

Davis, Jeffrey A.↗
NASA NTRS1990

Parallel-vector computation for CSI-design code

Computational aspects of Control-Structure Interaction (CSI) DESIGN code is reviewed. Numerical intensive computation portions of CSI-DESIGN code were identified. Improvements in computational speed for the CSI-DESIGN code can be achieved by exploiting parallel and vector capabilities offered by modern computers, such as the Alliant, Convex, Cray-2, and Cray-YMP. Four options to generate the coefficient stiffness matrix and to solve the system of linear, simultaneous equations are currently available in the CSI-DESIGN code. A preprocessor to use RCM (Reverse Cuthill-Mackee) algorithm for bandwidth minimization was also developed for the CSI-DESIGN code. Preliminary results obtained by solving a small-scale, 97 node CSI finite element model (for eigensolution) have indicated that this new CSI-DESIGN code is 5 to 6 times faster (using 1 Alliant processor) than the old version of CSI-DESIGN code. This speed-up was achieved due to the RCM algorithm and the use of a new skyline solver. Efforts are underway to further improve the vector speed for CSI-DESIGN code, to evaluate its performance on a larger scale CSI model (such as phase zero CSI model) to make the code run efficiently on multiprocessor, parallel computer environment, and to make the code portable among different parallel computers available at NASA LaRC, such as Alliant, Convex, and Cray computers.

Nguyen, Duc T.↗
NASA NTRS1973

Study of a hybrid multispectral processor

A hybrid processor is described offering enough handling capacity and speed to process efficiently the large quantities of multispectral data that can be gathered by scanner systems such as MSDS, SKYLAB, ERTS, and ERIM M-7. Combinations of general-purpose and special-purpose hybrid computers were examined to include both analog and digital types as well as all-digital configurations. The current trend toward lower costs for medium-scale digital circuitry suggests that the all-digital approach may offer the better solution within the time frame of the next few years. The study recommends and defines such a hybrid digital computing system in which both special-purpose and general-purpose digital computers would be employed. The tasks of recognizing surface objects would be performed in a parallel, pipeline digital system while the tasks of control and monitoring would be handled by a medium-scale minicomputer system. A program to design and construct a small, prototype, all-digital system has been started.

Marshall, R. E.↗
DOE OSTI2017

GPU-accelerated Tersoff potentials for massively parallel Molecular Dynamics simulations

The Tersoff potential is one of the empirical many-body potentials that has been widely used in simulation studies at atomic scales. Unlike pair-wise potentials, the Tersoff potential involves three-body terms, which require much more arithmetic operations and data dependency. In this contribution, we have implemented the GPU-accelerated version of several variants of the Tersoff potential for LAMMPS, an open-source massively parallel Molecular Dynamics code. Compared to the existing MPI implementation in LAMMPS, the GPU implementation exhibits a better scalability and offers a speedup of 2.2X when run on 1000 compute nodes on the Titan supercomputer. On a single node, the speedup ranges from 2.0 to 8.0 times, depending on the number of atoms per GPU and hardware configurations. The most notable features of our GPU-accelerated version include its design for MPI/accelerator heterogeneous parallelism, its compatibility with other functionalities in LAMMPS, its ability to give deterministic results and to support both NVIDIA CUDA- and OpenCL-enabled accelerators. Our implementation is now part of the GPU package in LAMMPS and accessible for public use.

Computer Science↗
DOE OSTI2019

Harnessing billions of tasks for a scalable portable hydrodynamic simulation of the merger of two stars

We propose a highly scalable demonstration of a portable asynchronous many-task programming model and runtime system applied to a grid-based adaptive mesh refinement hydrodynamic simulation of a double white dwarf merger with 14 levels of refinement that spans 17 orders of magnitude in astrophysical densities. The code uses the portable C++ parallel programming model that is embodied in the HPX library and being incorporated into the ISO C++ standard. The model reflects a significant shift from existing bulk synchronous parallel programming models under consideration for exascale systems. Through the use of the Futurization technique, seemingly sequential code is transformed into wait-free asynchronous tasks. We demonstrate the potential of our model by showing results from strong scaling runs on National Energy Research Scientific Computing Center’s Cori system (658,784 Intel Knight’s Landing cores) that achieve a parallel efficiency of 96.8% using billions of asynchronous tasks.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2020

Sage: parallel semi-asymmtric graph algorithms for NVRAMs

Non-volatile main memory (NVRAM) technologies provide an attractive set of features for large-scale graph analytics, including byte-addressability, low idle power, and improved memory-density. NVRAM systems today have an order of magnitude more NVRAM than traditional memory (DRAM). NVRAM systems could therefore potentially allow very large graph problems to be solved on a single machine, at a modest cost. However, a significant challenge in achieving high performance is in accounting for the fact that NVRAM writes can be much more expensive than NVRAM reads. In this paper, we propose an approach to parallel graph analytics using the Parallel Semi-Asymmetric Model (PSAM), in which the graph is stored as a read-only data structure (in NVRAM), and the amount of mutable memory is kept proportional to the number of vertices. Similar to the popular semi-external and semi-streaming models for graph analytics, the PSAM approach assumes that the vertices of the graph fit in a fast read-write memory (DRAM), but the edges do not. In NVRAM systems, our approach eliminates writes to the NVRAM, among other benefits. To experimentally study this new setting, we develop Sage, a parallel semi-asymmetric graph engine with which we implement provably-efficient (and often work-optimal) PSAM algorithms for over a dozen fundamental graph problems. We experimentally study Sage using a 48--core machine on the largest publicly-available real-world graph (the Hyperlink Web graph with over 3.5 billion vertices and 128 billion edges) equipped with Optane DC Persistent Memory, and show that Sage outperforms the fastest prior systems designed for NVRAM. Importantly, we also show that Sage nearly matches the fastest prior systems running solely in DRAM, by effectively hiding the costs of repeatedly accessing NVRAM versus DRAM.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2022

Enabling machine learning-ready HPC ensembles with Merlin

With the growing complexity of computational and experimental facilities, many scientific researchers are turning to machine learning (ML) techniques to analyze large scale ensemble data. With complexities such as multi-component workflows, heterogeneous machine architectures, parallel file systems, and batch scheduling, care must be taken to facilitate this analysis in a high performance computing (HPC) environment. Here, we present Merlin, a workflow framework to enable large ML-friendly ensembles of scientific HPC simulations. By augmenting traditional HPC with distributed compute technologies, Merlin aims to lower the barrier for scientific subject matter experts to incorporate ML into their analysis. As a producer–consumer workflow model, Merlin enables multi-machine, cross-batch job, dynamically allocated yet persistent workflows capable of utilizing surge-compute resources. Key features of Merlin are a flexible HPC-centric interface, low per-task overhead, multi-tiered fault recovery, and a hierarchical sampling algorithm that allows for $\mathscr{O}$(N) task execution and $\mathscr{O}$(N ln N) task queuing to ensembles of millions of tasks. In addition to Merlin’s design, we test the algorithm’s performance in an HPC center and demonstrate the ability to enqueue 40 million simulations in 100 s, with a 30 millisecond per-task overhead that is independent of ensemble size. Finally, we describe some example applications that Merlin has enabled on leadership-class HPC resources, such as the ML-augmented optimization of nuclear fusion experiments and the calibration of infectious disease models to study the progression of and possible mitigation strategies for COVID-19.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2017

Ab Initio Reactive Computer Aided Molecular Design

Few would dispute that theoretical chemistry tools can now provide keen insights into chemical phenomena. Yet the holy grail of efficient and reliable prediction of complex reactivity has remained elusive. Fortunately, recent advances in electronic structure theory based on the concepts of both element- and rank-sparsity, coupled with the emergence of new highly parallel computer architectures, have led to a significant increase in the time and length scales which can be simulated using first principles molecular dynamics. This then opens the possibility of new discovery-based approaches to chemical reactivity, such as the recently proposed ab initio nanoreactor. Here, we argue that due to these and other recent advances, the holy grail of computational discovery for complex chemical reactivity is rapidly coming within our reach.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2019

Parallelizable microfluidic dropmakers with multilayer geometry for the generation of double emulsions

Microfluidic devices enable the production of uniform double emulsions with control over droplet size and shell thickness. However, the limited production rate of microfluidic devices precludes the use of monodisperse double emulsions for industrial-scale applications, which require large quantities of droplets. To increase throughput, devices can be parallelized to contain many dropmakers operating simultaneously in one chip, but this is challenging to do for double emulsion dropmakers. Production of double emulsions requires dropmakers to have both hydrophobic and hydrophilic channels, requiring spatially precise patterning of channel surface wettability. Precise wettability patterning is difficult for devices containing multiple dropmakers, posing a significant challenge for parallelization. In this paper, we present a multilayer dropmaker geometry that greatly simplifies the process of producing microfluidic devices with excellent spatial control over channel wettability. Wettability patterning is achieved through the independent functionalization of channels in each layer prior to device assembly, rendering the dropmaker with a precise step between hydrophobic and hydrophilic channels. We report this device geometry enables uniform wettability patterning of parallelized dropmakers, providing a scalable approach for the production of double emulsions.

36 MATERIALS SCIENCE↗
DOE OSTI2020

Optically tunable mesoscale CdSe morphologies via inorganic phototropic growth

Inorganic phototropic growth using only spatially conformal illumination generated Se–Cd films that exhibited precise light-defined mesoscale morphologies including highly ordered, anisotropic, and periodic ridge and trench nanotextures over entire macroscopic substrates. Growth was accomplished via a light-induced electrochemical method using an optically and chemically isotropic solution, an unpatterned substrate, and unstructured, incoherent, low-intensity illumination in the absence of chemical directing agents or physical templates and masks. The morphologies were defined by the illumination inputs: the nanotexture long axes aligned parallel to the optical E-field vector, and the feature sizes and periods scaled with the wavelength. Optically based modeling of the growth closely reproduced the experimental results, confirming the film morphologies were fully determined by the light–matter interactions during growth. Solution processing of the Se–Cd films resulted in stoichiometric, crystalline CdSe films that also exhibited ordered nanotextures, demonstrating that inorganic phototropic growth can effect tunable, template-free generation of ordered CdSe nanostructures over macroscopic length scales.

36 MATERIALS SCIENCE↗
DOE OSTI2018

The reduction of magnetic reconnection outflow jets to sub-Alfvénic speeds

The outflow velocity of jets produced by collisionless magnetic reconnection is shown to be reduced by the ion exhaust temperature in fully kinetic particle in cell simulations and in situ satellite observations. We derive a scaling relationship for the outflow velocity based on the upstream Alfvén speed and the parallel ion exhaust temperature, which is verified in kinetic simulations and observations. The outflow speed reduction is shown to be due to the firehose instability criterion, and so, for large enough guide fields, this effect is suppressed and the outflow speed reaches the upstream Alfvén speed based on the reconnecting component of the magnetic field.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗