Search NASA⌕ Search

SEARCH · Search NASA

Results for “distributed 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 577 records · Page 32

Implementing efficient data compression and encryption in a persistent key-value store for HPC

Recently, persistent data structures, like key-value stores (KVSs), which are stored in a high-performance computing (HPC) system’s nonvolatile memory, provide an attractive solution for a number of emerging challenges like limited I/O performance. Data compression and encryption are two well-known techniques for improving several properties of such data-oriented systems. This article investigates how to efficiently integrate data compression and encryption into persistent KVSs for HPC with the ultimate goal of hiding their costs and complexity in terms of performance and ease of use. Our compression technique exploits deep memory hierarchy in an HPC system to achieve both storage reduction and performance improvement. Our encryption technique provides a practical level of security and enables sharing of sensitive data securely in complex scientific workflows with nearly imperceptible cost. We implement the proposed techniques on top of a distributed embedded KVS to evaluate the benefits and costs of incorporating these capabilities along different points in the dataflow path, illustrating differences in effective bandwidth, latency, and additional computational expense on Swiss National Supercomputing Centre’s Grand Tavé and National Energy Research Scientific Computing Center’s Cori.

97 MATHEMATICS AND COMPUTING↗

Concurrent Monte Carlo transport and fluence optimization with fluence adjusting scalable transport Monte Carlo

Purpose: The future of radiation therapy will require advanced inverse planning solutions to support single-arc, multiple-arc, and “4π” delivery modes, which present unique challenges in finding an optimal treatment plan over a vast search space, while still preserving dosimetric accuracy. The successful clinical implementation of such methods would benefit from Monte Carlo (MC) based dose calculation methods, which can offer improvements in dosimetric accuracy when compared to deterministic methods. The standard method for MC based treatment planning optimization leverages the accuracy of the MC dose calculation and efficiency of well-developed optimization methods, by precalculating the fluence to dose relationship within a patient with MC methods and subsequently optimizing the fluence weights. However, the sequential nature of this implementation is computationally time consuming and memory intensive. Methods to reduce the overhead of the MC precalculation have been explored in the past, demonstrating promising reductions of computational time overhead, but with limited impact on the memory overhead due to the sequential nature of the dose calculation and fluence optimization. The authors propose an entirely new form of “concurrent” Monte Carlo treat plan optimization: a platform which optimizes the fluence during the dose calculation, reduces wasted computation time being spent on beamlets that weakly contribute to the final dose distribution, and requires only a low memory footprint to function. In this initial investigation, the authors explore the key theoretical and practical considerations of optimizing fluence in such a manner. Methods: The authors present a novel derivation and implementation of a gradient descent algorithm that allows for optimization during MC particle transport, based on highly stochastic information generated through particle transport of very few histories. A gradient rescaling and renormalization algorithm, and the concept of momentum from stochastic gradient descent were used to address obstacles unique to performing gradient descent fluence optimization during MC particle transport. The authors have applied their method to two simple geometrical phantoms, and one clinical patient geometry to examine the capability of this platform to generate conformal plans as well as assess its computational scaling and efficiency, respectively. Results: The authors obtain a reduction of at least 50% in total histories transported in their investigation compared to a theoretical unweighted beamlet calculation and subsequent fluence optimization method, and observe a roughly fixed optimization time overhead consisting of ∼10% of the total computation time in all cases. Finally, the authors demonstrate a negligible increase in memory overhead of ∼7–8 MB to allow for optimization of a clinical patient geometry surrounded by 36 beams using their platform. Conclusions: This study demonstrates a fluence optimization approach, which could significantly improve the development of next generation radiation therapy solutions while incurring minimal additional computational overhead.

60 APPLIED LIFE SCIENCES↗

Modular Spiking Neural Circuits for Mapping Long Short-Term Memory on a Neurosynaptic Processor

Due to the distributed and asynchronous nature of neural computation through low-energy spikes, brain-inspired hardware systems offer high energy efficiency and massive parallelism. One such platform is the IBM TrueNorth neurosynaptic system. Recently, TrueNorth compatible representation learning algorithms have emerged, achieving close to the state-of-the-art performance in various data sets. However, its application in temporal sequence processing models, such as recurrent neural networks (RNNs), is still only at the proof of concept level. There is an inherent difficulty in capturing temporal dynamics of an RNN using spiking neurons, which is only exasperated by the hardware constraints in connectivity and synaptic weight resolution. This paper presents a design flow that overcomes these difficulties and maps a special case of recurrent networks called long short-term memory (LSTM) onto a spike-based platform. The framework utilizes various approximation techniques, such as activation discretization, weight quantization, and scaling and rounding, spiking neural circuits that implement the complex gating mechanisms, and a store-and-release technique to enable neuron synchronization and faithful storage. While the presented techniques can be applied to map LSTM to any spiking neural network (SNN) simulator/emulator, here we choose the TrueNorth chip as the target platform by adhering to its hardware constraints. Three LSTM applications, parity check, extended Reber grammar, and question classification, are evaluated. The tradeoffs among accuracy, performance, and energy tradeoffs achieved on TrueNorth are demonstrated. Finally, this is compared with the performance on an SNN platform without hardware constraints, which represents the upper bound of the achievable accuracy.

97 MATHEMATICS AND COMPUTING↗

A generalized massively parallel ultra-high order FFT-based Maxwell solver

Dispersion-free ultra-high order FFT-based Maxwell solvers have recently proven to be paramount to a large range of applications, including the high-fidelity modeling of high-intensity laser–matter interactions with Particle-In-Cell (PIC) codes. To enable a massively parallel scaling of these solvers, a novel parallelization technique was recently proposed, which consists in splitting the simulation domain into several processor sub-domains, with guard regions appended at each sub-domain boundary. Maxwell's equations are advanced independently on each sub-domain using local shared-memory FFTs (instead of a single distributed global FFT). This implies small truncation errors at sub-domain boundaries, the amplitude of which depends on guard regions sizes and order of the Maxwell solver. For moderate guard region sizes, this ’local’ technique proved to be highly scalable on up to a million cores and notably enabled the 3D modeling of so-called plasma mirrors, for which 8 guard cells only were enough to prevent truncation error growth. Yet, for other applications, the required number of guard cells might be much higher, which would severely limit the parallel efficiency of this technique due to the large volume of guard cells to be exchanged between sub-domains. In this context, we propose a novel parallelization technique that ensures very good scaling of FFT-based solvers with an arbitrarily high number of guard cells. Our ’hybrid’ technique consists in performing distributed FFTs on local groups of processors with guard regions now appended to boundaries of each group of processors. It uses a dual domain decomposition method for the Maxwell solver and other parts of the PIC cycle to keep the simulation load-balanced. This ’hybrid’ technique was implemented in the open source exascale library PICSAR. Benchmarks show that for a large number of guard cells (>16), the ’hybrid’ technique offers up to ×3 speed-up and ×8 memory savings compared to the ’local’ one.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Distribution of nanoscale nuclei in the amorphous dome of a phase change random access memory

The nanoscale crystal nuclei in an amorphous Ge{sub 2}Sb{sub 2}Te{sub 5} bit in a phase change memory device were evaluated by fluctuation transmission electron microscopy. The quench time in the device (∼10 ns) afforded more and larger nuclei in the melt-quenched state than in the as-deposited state. However, nuclei were even more numerous and larger in a test structure with a longer quench time (∼100 ns), verifying the prediction of nucleation theory that slower cooling produces more nuclei. It also demonstrates that the thermal design of devices will strongly influence the population of nuclei, and thus the speed and data retention characteristics.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗

A simple and efficient random walk solution of multi-rate mobile/immobile mass transport equations

We extend the particle-tracking method to simulate general multi-rate mass transfer (MRMT) equations. Previous methods for single-rate equations used two-state Markov chains and found that the time a par- ticle spends in the mobile state between waiting time epochs is random and exponentially distributed. Using Bochner's subordination technique for Markov processes, we find that the random mobile times are still exponential for the stochastic process that corresponds to the MRMT equations. The random times in the immobile phase have a distribution that is directly related to the memory function of the MRMT equation. This connection allows us to interpret the MRMT memory function as the rate at which particles of a certain age, measured by residence time in the immobile zone, exit to become mobile once again. Because the exact distributions of mobile and immobile times are known from the MRMT equa- tions, they can be simulated very simply and efficiently using random walks.

58 GEOSCIENCES↗

Experimental investigation of the residues produced in the {sup 136}Xe+Pb and {sup 124}Xe+Pb fragmentation reactions at 1A GeV

The nuclide cross sections and longitudinal velocity distributions of residues produced in the reactions of {sup 136}Xe and {sup 124}Xe at 1A GeV in a lead target were measured at the high-resolution magnetic spectrometer, the fragment separator (FRS) of GSI. The data cover a broad range of isotopes of the elements between Z=3 and Z=56 for {sup 136}Xe and between Z=5 and Z=55 for {sup 124}Xe, reaching down to cross sections of a few microbarns. The velocity distributions exhibit a Gaussian shape for masses above A=20, while more complex behavior is observed for lighter masses. The isotopic distributions for both reactions preserve a memory on the projectile N/Z ratio over the whole residue mass range.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

Dark matter transfer function: Free streaming, particle statistics, and memory of gravitational clustering

The transfer function T(k) of dark matter (DM) perturbations during matter domination is obtained by solving the linearized collisionless Boltzmann-Vlasov equation. We provide an exact expression for T(k) for arbitrary distribution functions of decoupled particles and initial conditions, which can be systematically expanded in a Fredholm series. An exhaustive numerical study of thermal relics for different initial conditions reveals that the first two terms in the expansion of T(k) provide a remarkably accurate and simple approximation valid on all scales of cosmological relevance for structure formation in the linear regime. The natural scale of suppression is the free-streaming wave vector at matter-radiation equality, k{sub fs}(t{sub eq})=[4{pi}{rho}{sub 0M}/[<V-vector{sup 2}>(1+z{sub eq})]]{sup 1/2}. An important ingredient is a nonlocal kernel determined by the distribution functions of the decoupled particles which describes the memory of the initial conditions and gravitational clustering and yields a correction to the fluid description. This correction is negligible at large scales k<<k{sub fs}(t{sub eq}) but it becomes important at small scales k{>=}k{sub fs}(t{sub eq}). Distribution functions that favor the small momentum region yield longer-range memory kernels and lead to an enhancement of power at small scales k>k{sub fs}(t{sub eq}). Fermi-Dirac and Bose-Einstein statistics lead to long-range memory kernels, with longer-range for bosons, both resulting in enhancement of T(k) at small scales. For DM thermal relics that decoupled while ultrarelativistic we find k{sub fs}(t{sub eq}){approx_equal}0.003(g{sub d}/2){sup 1/3} (m/keV) [kpc]{sup -1}, where g{sub d} is the number of degrees of freedom at decoupling. For WIMPS we obtain k{sub fs}(t{sub eq})=5.88(g{sub d}/2){sup 1/3} (m/100 GeV){sup 1/2} (T{sub d}/10 MeV){sup 1/2} [pc]{sup -1}. For k<<k{sub fs}(t{sub eq}), T(k){approx}1-C[k/k{sub fs}(t{sub eq})]{sup 2} where C=O(1) and independent of statistics for thermal relics. We provide simple and accurate fits for T(k) in a wide range of small scales k>k{sub fs}(t{sub eq}) for thermal relics and different initial conditions. The numerical and analytic results for arbitrary distribution functions and initial conditions allow an assessment of DM candidates through their impact on structure formation.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Multiplexed color centers in a silicon photonic cavity array

Entanglement distribution is central to the modular scaling of quantum processors and establishing quantum networks. Color centers with telecom-band transitions and long spin coherence times are suitable candidates for long-distance entanglement distribution. However, high-bandwidth memory-enhanced quantum communication is limited by high-yield, scalable creation of efficient spin-photon interfaces. Here, we develop a silicon photonics platform consisting of arrays of bus-coupled cavities. The coupling to a common bus waveguide enables simultaneous access to individually addressable cavity-enhanced T center arrays. We demonstrate frequency-multiplexed operation of two T centers in separate photonic crystal cavities. In addition, we investigate the cavity enhancement of a T center through hybridized modes formed between physically distant cavities. Our results show that bus-coupled arrays of cavity-enhanced color centers could enable efficient on-chip and long-distance entanglement distribution.

Komza, Lukasz↗

Clustering properties of a sterile neutrino dark matter candidate

The clustering properties of sterile neutrinos are studied within a simple extension of the minimal standard model, where these neutrinos are produced via the decay of a gauge singlet scalar. The distribution function after decoupling is strongly out of equilibrium and features an enhancement at small comoving momentum {proportional_to}1/{radical}(p). Dark matter abundance and phase space density constraints from dwarf spheroidal galaxies constrain the mass in the keV range consistent with a Yukawa coupling to a gauge singlet with mass and vacuum expectation value in the range {approx}100 GeV and a decoupling temperature of this order. The dark matter transfer function and power spectrum are obtained from the solution of the nonrelativistic Boltzmann-Vlasov equation in the matter dominated era. The small momentum enhancement of the nonequilibrium distribution function leads to long range memory of gravitational clustering and a substantial enhancement of the power spectrum at small scales as compared to a thermal relic or sterile neutrino produced via nonresonant mixing with active neutrinos. The scale of suppression of the power spectrum for a sterile neutrino with m{approx}keV produced by scalar decay that decouples at {approx}100 GeV is {lambda}{approx}488 kpc. At large scales T(k){approx}1-Ck{sup 2}/k{sub fs}{sup 2}(t{sub eq})+{center_dot}{center_dot}{center_dot} with C{approx}O(1). At small scales 65 kpc < or approx. {lambda} < or approx. 500 kpc corrections to the fluid description and memory of gravitational clustering become important, and we find T(k){approx_equal}1.902e{sup -k/k{sub fs}(t{sub eq})}, where k{sub fs}(t{sub eq}){approx}0.013/kpc is the free-streaming wave vector at matter-radiation equality. The enhancement of power at small scales may provide possible relief to the tension between the constraints from x-ray and Lyman-{alpha} forest data.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Extremely Scalable Distributed Computation of Contour Trees via Pre-Simplification

Contour trees offer an abstract representation of the level set topology in scalar fields and are widely used in topological data analysis and visualization. However, applying contour trees to large-scale scientific datasets remains challenging due to scalability limitations. Recent developments in distributed hierarchical contour trees have addressed these challenges by enabling scalable computation across distributed systems. Building on these structures, advanced analytical tasks—such as volumetric branch decomposition and contour extraction—have been introduced to facilitate large-scale scientific analysis. Despite these advancements, such analytical tasks substantially increase memory usage, which hampers scalability. In this paper, we propose a pre-simplification strategy to significantly reduce the memory overhead associated with analytical tasks on distributed hierarchical contour trees. We demonstrate enhanced scalability through strong scaling experiments, constructing the largest known contour tree—comprising over half a trillion nodes with complex topology—in under 15 minutes on a dataset containing 550 billion elements.

Li, Mingzhe [University of Utah]↗

Scalable Transparent Checkpoint-Restart of Global Address Space Applications on Virtual Machines over Infiniband

Checkpoint-Restart is one of the most used software approaches to achieve fault-tolerance in high-end clusters. While standard techniques typically focus on user-level solutions, the advent of virtualization software has enabled efficient and transparent system-level approaches. In this paper, we present a scalable transparent system-level solution to address fault-tolerance for applications based on global address space (GAS) programming models on Infiniband clusters. In addition to handling communication, the solution addresses transparent checkpoint of user-generated files. We exploit the support for the Infiniband network in the Xen virtual machine environment. We have developed a version of the Aggregate Remote Memory Copy Interface (ARMCI) one-sided communication library capable of suspending and resuming applications. We present efficient and scalable mechanisms to distribute checkpoint requests and to backup virtual machines memory images and file systems. We tested our approach in the context of NWChem, a popular computational chemistry suite. We demonstrated that NWChem can be executed, without any modification to the source code, on a virtualized 8-node cluster with very little overhead (below 3%). We observe that the total checkpoint time is limited by disk I/O. Finally, we measured system-size depended components of the checkpoint time on up to 1024 cores (128 nodes), demonstrating the scalability of our approach in medium/large-scale systems.

97 MATHEMATICS AND COMPUTING↗

UNITY: Unified Memory and Storage Space

UNITY is a 36-month project focused on providing design and evaluate a new distributed storage paradigm that unifies the traditionally distinct application views of memory- and file-based data storage into a single scalable and resilient environment. The project is a collaboration among Oak Ridge National Laboratory (ORNL, Lead institution), Los Alamos National Labs (LANL), and Georgia Tech (GT). The main contributions of the GT team have been around development of low level systems software for best leveraging the capabilities of new types of persistent memory technologies, and for development of methods for intelligent data management across memory/storage substrates with heterogeneous components. GT contributed the Phoenix library for optimized checkpoint/restart for HPC I/O for systems with non-volatile memory (NVM), the NVStream library for NVM-specialized streaming I/O for HPC workflows, the CoMerge, Mnemo and Kleio solutions for intelligent data management on NVM-based systems. These contributions result in significant improvements in both application performance and system efficiency.

97 MATHEMATICS AND COMPUTING↗

Authenticating cache

The aim of the Authenticating Cache architecture is to ensure that machine instructions in a Read Only Memory (ROM) are legitimate from the time the ROM image is signed (immediately after compilation) to the time they are placed in the cache for the processor to consume. The proposed architecture allows the detection of ROM image modifications during distribution or when it is loaded into memory. It also ensures that modified instructions will not execute in the processor-as the cache will not be loaded with a page that fails an integrity check. The authenticity of the instruction stream can also be verified in this architecture. The combination of integrity and authenticity assurance greatly improves the security profile of a system.

97 MATHEMATICS AND COMPUTING↗

Persistent Octrees for Parallel Mesh Refinement Through Non-Volatile Byte-Addressable Memory

We report that octree-based mesh adaptation has enabled simulations of complex physical phenomena. Existing meshing algorithms were proposed with the assumption that computer memory is volatile. Consequently, for failure recovery, the in-core algorithms need to save memory states as snapshots with slow file I/Os. The out-of-core algorithms store octants on disks for persistence. However, neither of them was designed to leverage unique characteristics of non-volatile byte-addressable memory (NVBM). In this paper, we propose a novel data structure Distributed Persistent Merged octree (DPM-octree) for both meshing and in-memory storage of persistent octrees using NVBM. It is a multi-version data structure and can recover from failures using its earlier persistent version stored in NVBM. In addition, we design a feature-directed sampling approach to help dynamically transform the DPM-octree layout for reducing NVBM-induced memory write latency. DPM-octree uses parity trees which are created using erasure coding and stored in NVBM to support low-latency in-memory octant recovery after data loss. DPM-octree has been successfully integrated with Gerris software for simulation of fluid dynamics. Finally, our experimental results with real-world scientific workloads show that DPM-octree scales up to 1.1 billion mesh elements with 1000 processors on the Titan supercomputer.

97 MATHEMATICS AND COMPUTING↗

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↗

A Communication-Optimal Framework for Contracting Distributed Tensors

Tensor contractions are extremely compute intensive generalized matrix multiplication operations encountered in many computational science fields, such as quantum chemistry and nuclear physics. Unlike distributed matrix multiplication, which has been extensively studied, limited work has been done in understanding distributed tensor contractions. In this paper, we characterize distributed tensor contraction algorithms on torus networks. We develop a framework with three fundamental communication operators to generate communication-efficient contraction algorithms for arbitrary tensor contractions. We show that for a given amount of memory per processor, our framework is communication optimal for all tensor contractions. We demonstrate performance and scalability of our framework on up to 262,144 cores of BG/Q supercomputer using five tensor contraction examples.

communication efficiency↗

The Development of a Generalized Riser Flow Regime Map Based Upon Higher Moment and Chaotic Statistics Using Electrical Capacitance Volume Tomography (ECVT)

Dynamic analyses have been applied to the temporal signals from an Electro Capacitance Volume Tomography instrument located near mid-height on the riser of an industrial-scale cold-flow circulating fluidized bed to characterize gas-solids flow behavior in the riser. Twelve capacitance electrodes surround the cylindrical riser over a height of 1.3 m. The instrument used a neural network deconvolution algorithm to determine the spatially resolved solids fraction recorded at 52 Hz. Experiments were carried out over a range of gas and solids flows in the transport regime using a Geldart Group B bed material, high density polyethylene with mean particle size of 880 μm. The radial solids distribution was found to vary from one-time step to the next between profiles typical of laminar and turbulent flow. The duration of time spent in each of these flow profiles depended upon the operating regime – dilute, core-annular, or fast fluidized bed. The chaotic structure of the temporal data was characterized using the three conventional approaches: the first 4 moments from the distribution of signal in time, system memory parameters from the autocorrelation function and the Hurst exponent, and analysis of the correlationentropy and correlation dimension of the attractor. These signal analysis techniques were used to clearly distinguish differences between different transport operating regimes. Specifically, it was experimentally observed that a riser transitions from core annular flow profile to dilute and dense regimes via increasing the frequency of short term transients to either dilute or dense flow profiles, respectively. A regime map was generated based upon these dynamics using solids flux and gas velocity axes. Fast fluidized, core annular, and dilute each exhibited different degree of dynamic characteristics typical of fluid dominated or particle compromising behavior. It should be noted that the magnitude for the different statistics was in the same range regardless of the regime, it was the radial profile for the statistic that changed and subsequently identified that there was a change in the regime. Finally, a reduced regime map was developed consisting of plotting the gas velocity normalized by the upper transport velocity versus the solids flux normalized by the saturation carrying capacity. The use of this reduced plot allowed the data from widely different conditions to be plotted and compared on the same<p>graph. Note that in many instances, some of the statistics identified the operating point as being in one regime while others indicated that it was in another indicating a transition region between dilute or core annular regimes and between the core annular and fast fluidization regimes. This now provides a tool that can be used to optimize process performance, identify changes in operating states, or replicate process dynamics during process scaling or changing operating parameters. </p>

Breault, Ronald↗