Search NASA⌕ Search

SEARCH · Search NASA

Results for “memory efficiency”

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 1,081 records · Page 60

Accurate, Fast and Scalable Kernel Ridge Regression on Parallel and Distributed Systems

Kernel Ridge Regression (KRR) is a fundamental method in machine learning. Given an n-by-d data matrix as input, a traditional implementation requires Θ(n2) memory to form an n-by-n kernel matrix and Θ(n3) flops to compute the final model. These time and storage costs prohibit KRR from scaling up to large datasets. For example, even on a relatively small dataset (a 520k-by-90 input requiring 357 MB), KRR requires 2 TB memory just to store the kernel matrix. Additionally, the reason is that n usually is much larger than d for real-world applications. On the other hand, weak scaling becomes a problem: if we keep d and n/p fixed as p grows (p is # machines), the memory needed grows as Θ(p) per processor and the flops as Θ(p2) per processor. In the perfect weak scaling situation, both the memory needed and the flops grow as Θ(1) per processor (i.e. memory and flops are constant). The traditional Distributed KRR implementation (DKRR) only achieved 0.32% weak scaling efficiency from 96 to 1536 processors. In this work, we propose two new methods to address these problems: the Balanced KRR (BKRR) and K-means KRR (KKRR). These methods consider alternative ways to partition the input dataset into p different parts, generating p different models, and then selecting the best model among them. Compared to a conventional implementation, KKRR2 (optimized version of KKRR) improves the weak scaling efficiency from 0.32% to 38% and achieves a 591x speedup for getting the same accuracy by using the same data and the same hardware (1536 processors). BKRR2 (optimized version of BKRR) achieves a higher accuracy than the current fastest method using less training time for a variety of datasets. For the applications requiring only approximate solutions, BKRR2 improves the weak scaling efficiency to 92% and achieves 3505x speedup (theoretical speedup: 4096x).

97 MATHEMATICS AND COMPUTING↗

Search for optimal distance spectrum convolutional codes

In order to communicate reliably and to reduce the required transmitter power, NASA uses coded communication systems on most of their deep space satellites and probes (e.g. Pioneer, Voyager, Galileo, and the TDRSS network). These communication systems use binary convolutional codes. Better codes make the system more reliable and require less transmitter power. However, there are no good construction techniques for convolutional codes. Thus, to find good convolutional codes requires an exhaustive search over the ensemble of all possible codes. In this paper, an efficient convolutional code search algorithm was implemented on an IBM RS6000 Model 580. The combination of algorithm efficiency and computational power enabled us to find, for the first time, the optimal rate 1/2, memory 14, convolutional code.

Connor, Matthew C.↗

Nonlinear structural response using adaptive dynamic relaxation on a massively-parallel-processing system

A parallel adaptive dynamic relaxation (ADR) algorithm has been developed for nonlinear structural analysis. This algorithm has minimal memory requirements, is easily parallelizable and scalable to many processors, and is generally very reliable and efficient for highly nonlinear problems. Performance evaluations on single-processor computers have shown that the ADR algorithm is reliable and highly vectorizable, and that it is competitive with direct solution methods for the highly nonlinear problems considered. The present algorithm is implemented on the 512-processor Intel Touchstone DELTA system at Caltech, and it is designed to minimize the extent and frequency of interprocessor communication. The algorithm has been used to solve for the nonlinear static response of two and three dimensional hyperelastic systems involving contact. Impressive relative speedups have been achieved and demonstrate the high scalability of the ADR algorithm. For the class of problems addressed, the ADR algorithm represents a very promising approach for parallel-vector processing.

Oakley, David R.↗

Wing-Body Aeroelasticity Using Finite-Difference Fluid/Finite-Element Structural Equations on Parallel Computers

In recent years significant advances have been made for parallel computers in both hardware and software. Now parallel computers have become viable tools in computational mechanics. Many application codes developed on conventional computers have been modified to benefit from parallel computers. Significant speedups in some areas have been achieved by parallel computations. For single-discipline use of both fluid dynamics and structural dynamics, computations have been made on wing-body configurations using parallel computers. However, only a limited amount of work has been completed in combining these two disciplines for multidisciplinary applications. The prime reason is the increased level of complication associated with a multidisciplinary approach. In this work, procedures to compute aeroelasticity on parallel computers using direct coupling of fluid and structural equations will be investigated for wing-body configurations. The parallel computer selected for computations is an Intel iPSC/860 computer which is a distributed-memory, multiple-instruction, multiple data (MIMD) computer with 128 processors. In this study, the computational efficiency issues of parallel integration of both fluid and structural equations will be investigated in detail. The fluid and structural domains will be modeled using finite-difference and finite-element approaches, respectively. Results from the parallel computer will be compared with those from the conventional computers using a single processor. This study will provide an efficient computational tool for the aeroelastic analysis of wing-body structures on MIMD type parallel computers.

Byun, Chansup↗

The Optimized Parallel MOC module in Code cosLATC

A new neutron transport lattice physics code, cosLATC, which is an essential part in COSINE (Core and System Integrated Engine) code package, has been developed by SPICRI. It is mainly used for 2-D transport calculations on Pressurized Water Reactor (PWR) assemblies or pin cells. cosLATC generates few group constants for the core physics code cosCORE in COSINE package. With regards to cross-section data, WIMS-D 69 and 281 energy groups library is employed currently. cosLATC also has capability of resonance treatment, depletion and fuel temperature calculation. cosLATC uses modular tracking and parallelization technique in MOC transport calculation, which can save lots of memory and running time. In this paper, MOC module in cosLATC are introduced, the accuracy and efficiency of the module are verified by C5G7 benchmark and a PWR typical assembly. (authors)

21 SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLANTS↗

Slave finite element for non-linear analysis of engine structures. Volume 2: Programmer's manual and user's manual

The programming aspects of SFENES are described in the User's Manual. The information presented is provided for the installation programmer. It is sufficient to fully describe the general program logic and required peripheral storage. All element generated data is stored externally to reduce required memory allocation. A separate section is devoted to the description of these files thereby permitting the optimization of Input/Output (I/O) time through efficient buffer descriptions. Individual subroutine descriptions are presented along with the complete Fortran source listings. A short description of the major control, computation, and I/O phases is included to aid in obtaining an overall familiarity with the program's components. Finally, a discussion of the suggested overlay structure which allows the program to execute with a reasonable amount of memory allocation is presented.

Witkop, D. L.↗

NASA Tech Briefs, November 2008

Topics covered include: Digital Phase Meter for a Laser Heterodyne Interferometer; Vision System Measures Motions of Robot and External Objects; Advanced Precipitation Radar Antenna to Measure Rainfall From Space; Wide-Band Radar for Measuring Thickness of Sea Ice; Vertical Isolation for Photodiodes in CMOS Imagers; Wide-Band Microwave Receivers Using Photonic Processing; L-Band Transmit/Receive Module for Phase-Stable Array Antennas; Microwave Power Combiner/Switch Utilizing a Faraday Rotator; Compact Low-Loss Planar Magic-T; Using Pipelined XNOR Logic to Reduce SEU Risks in State Machines; Quasi-Optical Transmission Line for 94-GHz Radar; Next Generation Flight Controller Trainer System; Converting from DDOR SASF to APF; Converting from CVF to AAF; Documenting AUTOGEN and APGEN Model Files; Sequence History Update Tool; Extraction and Analysis of Display Data; MRO DKF Post-Processing Tool; Rig Diagnostic Tools; MRO Sequence Checking Tool; Science Activity Planner for the MER Mission; UAVSAR Flight-Planning System; Templates for Deposition of Microscopic Pointed Structures; Adjustable Membrane Mirrors Incorporating G-Elastomers; Hall-Effect Thruster Utilizing Bismuth as Propellant; High-Temperature Crystal-Growth Cartridge Tubes Made by VPS; Quench Crucibles Reinforced with Metal; Deep-Sea Hydrothermal-Vent Sampler; Mars Rocket Propulsion System; Two-Stage Passive Vibration Isolator; Improved Thermal Design of a Compression Mold; Enhanced Pseudo-Waypoint Guidance for Spacecraft Maneuvers; Altimetry Using GPS-Reflection/Occultation Interferometry; Thermally Driven Josephson Effect; Perturbation Effects on a Supercritical C7H16/N2 Mixing Layer; Gold Nanoparticle Labels Amplify Ellipsometric Signals; Phase Matching of Diverse Modes in a WGM Resonator; WGM Resonators for Terahertz-to-Optical Frequency Conversion; Determining Concentration of Nanoparticles from Ellipsometry; Microwave-to-Optical Conversion in WGM Resonators; Four-Pass Coupler for Laser-Diode-Pumped Solid-State Laser; Low-Resolution Raman-Spectroscopy Combustion Thermometry; Temperature Sensors Based on WGM Optical Resonators; Varying the Divergence of Multiple Parallel Laser Beams; Efficient Algorithm for Rectangular Spiral Search; Algorithm-Based Fault Tolerance Integrated with Replication; Targeting and Localization for Mars Rover Operations; Terrain-Adaptive Navigation Architecture; Self-Adjusting Hash Tables for Embedded Flight Applications; Schema for Spacecraft-Command Dictionary; Combined GMSK Communications and PN Ranging; System-Level Integration of Mass Memory; Network-Attached Solid-State Recorder Architecture; Method of Cross-Linking Aerogels Using a One-Pot Reaction Scheme; An Efficient Reachability Analysis Algorithm.

Source record↗

A DRAM compiler algorithm for high performance VLSI embedded memories

In many applications, the limited density of the embedded SRAM does not allow integrating the memory on the same chip with other logic and functional blocks. In such cases, the embedded DRAM provides the optimum combination of very high density, low power, and high performance. For ASIC's to take full advantage of this design strategy, an efficient and highly reliable DRAM compiler must be used. The embedded DRAM architecture, cell, and peripheral circuit design considerations and the algorithm of a high performance memory compiler are presented .

Eldin, A. G.↗

Scalable Cloning on Large-Scale GPU Platforms with Application to Time-Stepped Simulations on Grids

Cloning is a technique to efficiently simulate a tree of multiple what-if scenarios that are unraveled during the course of a base simulation. However, cloned execution is highly challenging to realize on large, distributed memory computing platforms, due to the dynamic nature of the computational load across clones, and due to the complex dependencies spanning the clone tree. In this paper, we present the conceptual simulation framework, algorithmic foundations, and runtime interface of CloneX, a new system we designed for scalable simulation cloning. It efficiently and dynamically creates whole logical copies of a dynamic tree of simulations across a large parallel system without full physical duplication of computation and memory. The performance of a prototype implementation executed on up to 1,024 graphical processing units of a supercomputing system has been evaluated with three benchmarks—heat diffusion, forest fire, and disease propagation models—delivering a speed up of over two orders of magnitude compared to replicated runs. Finally, the results demonstrate a significantly faster and scalable way to execute many what-if scenario ensembles of large simulations via cloning using the CloneX interface.

97 MATHEMATICS AND COMPUTING↗

Memory monitoring by animals and humans

The authors asked whether animals and humans would use similarly an uncertain response to escape indeterminate memories. Monkeys and humans performed serial probe recognition tasks that produced differential memory difficulty across serial positions (e.g., primacy and recency effects). Participants were given an escape option that let them avoid any trials they wished and receive a hint to the trial's answer. Across species, across tasks, and even across conspecifics with sharper or duller memories, monkeys and humans used the escape option selectively when more indeterminate memory traces were probed. Their pattern of escaping always mirrored the pattern of their primary memory performance across serial positions. Signal-detection analyses confirm the similarity of the animals' and humans' performances. Optimality analyses assess their efficiency. Several aspects of monkeys' performance suggest the cognitive sophistication of their decisions to escape.

NASA Discipline Space Human Factors↗

Segmented Hexagonal Antenna Reflector Concentrically Stacked Using Shape Memory Composite Tubular Hinges

A new architecture for solid surface reflector antennas scalable to sizes greater than 10 m is presented. The design combines the compact, lightness and simplicity characteristics of advanced deployable structures to create sub-reflectors that can be assembled in space into larger units using a robotic arm and operations under development by a related parallel project. The seven-panel hexagonal sub-reflector is divided into hexagonal panels of that efficiently stack concentrically and vertically. The central panel is connected to each side panel on the back side by a pair of tubular shape memory composite hinges that enable the required deployment kinematics. A secondary mechanism ultimately closes the interpanel gap. The focus of the paper is on the development of the sub-reflector elements, namely the tubular hinges that use embedded heaters and sensors for actuation, the actuation mechanisms, and the lightweight sandwich construction panels. A finite element analysis parametric study assessed how design features of the hinge affect its stowage and deployment dynamics. The component fabrication and testing plans for the two-panel assembly breadboard model are outlined.

deployable reflectors↗

Segmented Hexagonal Antenna Reflector Concentrically Stacked Using Shape Memory Composite Tubular Hinges

A new architecture for solid surface reflector antennas scalable to sizes greater than 10 m is presented. The design combines the compact, lightness and simplicity characteristics of advanced deployable structures to create sub-reflectors that can be assembled in space into larger units using a robotic arm and operations under development by a related parallel project. The seven-panel hexagonal sub-reflector is divided into hexagonal panels of that efficiently stack concentrically and vertically. The central panel is connected to each side panel on the back side by a pair of tubular shape memory composite hinges that enable the required deployment kinematics. A secondary mechanism ultimately closes the interpanel gap. The focus of the paper is on the development of the sub-reflector elements, namely the tubular hinges that use embedded heaters and sensors for actuation, the actuation mechanisms, and the lightweight sandwich construction panels. A finite element analysis parametric study assessed how design features of the hinge affect its stowage and deployment dynamics. The component fabrication and testing plans for the two-panel assembly breadboard model are outlined.

deployable reflectors↗

Numerical integration of the extended variable generalized Langevin equation with a positive Prony representable memory kernel

Generalized Langevin dynamics (GLD) arise in the modeling of a number of systems, ranging from structured fluids that exhibit a viscoelastic mechanical response, to biological systems, and other media that exhibit anomalous diffusive phenomena. Molecular dynamics (MD) simulations that include GLD in conjunction with external and/or pairwise forces require the development of numerical integrators that are efficient, stable, and have known convergence properties. In this article, we derive a family of extended variable integrators for the Generalized Langevin equation with a positive Prony series memory kernel. Using stability and error analysis, we identify a superlative choice of parameters and implement the corresponding numerical algorithm in the LAMMPS MD software package. Salient features of the algorithm include exact conservation of the first and second moments of the equilibrium velocity distribution in some important cases, stable behavior in the limit of conventional Langevin dynamics, and the use of a convolution-free formalism that obviates the need for explicit storage of the time history of particle velocities. Capability is demonstrated with respect to accuracy in numerous canonical examples, stability in certain limits, and an exemplary application in which the effect of a harmonic confining potential is mapped onto a memory kernel.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

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 peripheral component interconnect express-based scalable and highly integrated pulsed spectrometer for solution state dynamic nuclear polarization

High sensitivity, high data rates, fast pulses, and accurate synchronization all represent challenges for modern nuclear magnetic resonance spectrometers, which make any expansion or adaptation of these devices to new techniques and experiments difficult. Here, we present a Peripheral Component Interconnect Express (PCIe)-based highly integrated distributed digital architecture pulsed spectrometer that is implemented with electron and nucleus double resonances and is scalable specifically for broad dynamic nuclear polarization (DNP) enhancement applications, including DNP-magnetic resonance spectroscopy/imaging (DNP-MRS/MRI). The distributed modularized architecture can implement more transceiver channels flexibly to meet a variety of MRS/MRI instrumentation needs. The proposed PCIe bus with high data rates can significantly improve data transmission efficiency and communication reliability and allow precise control of pulse sequences. An external high speed double data rate memory chip is used to store acquired data and pulse sequence elements, which greatly accelerates the execution of the pulse sequence, reduces the TR (time of repetition) interval, and improves the accuracy of TR in imaging sequences. Using clock phase-shift technology, we can produce digital pulses accurately with high timing resolution of 1 ns and narrow widths of 4 ns to control the microwave pulses required by pulsed DNP and ensure overall system synchronization. The proposed spectrometer is proved to be both feasible and reliable by observation of a maximum signal enhancement factor of approximately −170 for {sup 1}H, and a high quality water image was successfully obtained by DNP-enhanced spin-echo {sup 1}H MRI at 0.35 T.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

GPU-enabled extreme-scale turbulence simulations: Fourier pseudo-spectral algorithms at the exascale using OpenMP offloading

Fourier pseudo-spectral methods for nonlinear partial differential equations are of wide interest in many areas of advanced computational science, including direct numerical simulation of three-dimensional (3-D) turbulence governed by the Navier-Stokes equations in fluid dynamics. This paper presents a new capability for simulating turbulence at a new record resolution up to 35 trillion grid points, on the world's first exascale computer, Frontier, comprising AMD MI250x GPUs with HPE's Slingshot interconnect and operated by the US Department of Energy's Oak Ridge Leadership Computing Facility (OLCF). Key programming strategies designed to take maximum advantage of the machine architecture involve performing almost all computations on the GPU which has the same memory capacity as the CPU, performing all-to-all communication among sets of parallel processes directly on the GPU, and targeting GPUs efficiently using OpenMP offloading for intensive number-crunching including 1-D Fast Fourier Transforms (FFT) performed using AMD ROCm library calls. With 99% of computing power on Frontier being on the GPU, leaving the CPU idle leads to a net performance gain via avoiding the overhead of data movement between host and device except when needed for some I/O purposes. Memory footprint including the size of communication buffers for MPI_ALLTOALL is managed carefully to maximize the largest problem size possible for a given node count. Detailed performance data including separate contributions from different categories of operations to the elapsed wall time per step are reported for five grid resolutions, from 2048 3 on a single node to 32768 3 on 4096 or 8192 nodes out of 9408 on the system. Both 1D and 2D domain decompositions which divide a 3D periodic domain into slabs and pencils respectively are implemented. The present code suite (labeled by the acronym GESTS, GPUs for Extreme Scale Turbulence Simulations) achieves a figure of merit (in grid points per second) exceeding goals set in the Center for Accelerated Application Readiness (CAAR) program for Frontier. The performance attained is highly favorable in both weak scaling and strong scaling, with notable departures only for 2048 3 where communication is entirely intra-node, and for 32768 3 , where a challenge due to small message sizes does arise. Communication performance is addressed further using a lightweight test code that performs all-to-all communication in a manner matching the full turbulence simulation code. Performance at large problem sizes is affected by both small message size due to high node counts as well as dragonfly network topology features on the machine, but is consistent with official expectations of sustained performance on Frontier. Overall, although not perfect, the scalability achieved at the extreme problem size of 32768 3 (and up to 8192 nodes — which corresponds to hardware rated at just under 1 exaflop/sec of theoretical peak computational performance) is arguably better than the scalability observed using prior state-of-the-art algorithms on Frontier's predecessor machine (Summit) at OLCF. New science results for the study of intermittency in turbulence enabled by this code and its extensions are to be reported separately in the near future.

3D fast Fourier transform↗

An efficient approach to unstructured mesh hydrodynamics on the cell broadband engine (u)

Unstructured mesh physics for the Cell Broadband Engine (CBE) has received little or no attention to date, largely because the CBE architecture poses particular challenges for unstructured mesh algorithms. SPU memory management strategies such as data preloading cannot be applied to the irregular memory storage patterns of unstructured meshes; and the SPU vector instruction set does not support the indirect addressing needed by connectivity arrays. This paper presents an approach to unstructured mesh physics that addresses these challenges, by creating a new mesh data structure and reorganizing code to give efficient CBE performance. The approach is demonstrated on the FLAG production hydrodynamics code using standard test problems, and results show an average speedup of more than 5x over the original code.

99 GENERAL AND MISCELLANEOUS↗