Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Multiple processors”

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 109 records · Page 6

Multi-fault Tolerance for Cartesian Data Distributions

Faults are expected to play an increasingly important role in how algorithms and applications are designed to run on future extreme-scale sys- tems. Algorithm-based fault tolerance (ABFT) is a promising approach that involves modications to the algorithm to recover from faults with lower over- heads than replicated storage and a signicant reduction in lost work compared to checkpoint-restart techniques. Fault-tolerant linear algebra (FTLA) algo- rithms employ additional processors that store parities along the dimensions of a matrix to tolerate multiple, simultaneous faults. Existing approaches as- sume regular data distributions (blocked or block-cyclic) with the failures of each data block being independent. To match the characteristics of failures on parallel computers, we extend these approaches to mapping parity blocks in several important ways. First, we handle parity computation for generalized Cartesian data distributions with each processor holding arbitrary subsets of blocks in a Cartesian-distributed array. Second, techniques to handle corre- lated failures, i.e., multiple processors that can be expected to fail together, are presented. Third, we handle the colocation of parity blocks with the data blocks and do not require them to be on additional processors. Several al- ternative approaches, based on graph matching, are presented that attempt to balance the memory overhead on processors while guaranteeing the same fault tolerance properties as existing approaches that assume independent fail- ures on regular blocked data distributions. The evaluation of these algorithms demonstrates that the additional desirable properties are provided by the pro- posed approach with minimal overhead.

Ali, Nawab↗

A Four Channel Beam Current Monitor Data Acquisition System Using Embedded Processors

Data acquisition from multiple beam current monitors is required for electron accelerator production of Mo-99. A two channel system capable of recording data from two beam current monitors has been developed, is currently in use, and is discussed below. The development of a cost-effective method of extending this system to more than two channels and integrating of these measurements into an accelerator control system is the main focus of this report. Data from these current monitors is digitized, processed, and stored by a digital data acquisition system. Limitations and drawbacks with the currently deployed digital data acquisition system have been identified as have been potential solutions, or at least improvements, to these problems. This report will discuss and document the efforts we've made in improving the flexibility and lowering the cost of the data acquisition system while maintaining the minimum requirements.

42 ENGINEERING↗

Deterministic multi-qubit entanglement in a quantum network

High-fidelity deterministic quantum state transfer and multi-qubit entanglement are demonstrated in a quantum network comprising two superconducting quantum nodes one metre apart, with each node including three interconnected qubits. The generation of high-fidelity distributed multi-qubit entanglement is a challenging task for large-scale quantum communication and computational networks. The deterministic entanglement of two remote qubits has recently been demonstrated with both photons and phonons. However, the deterministic generation and transmission of multi-qubit entanglement has not been demonstrated, primarily owing to limited state-transfer fidelities. Here we report a quantum network comprising two superconducting quantum nodes connected by a one-metre-long superconducting coaxial cable, where each node includes three interconnected qubits. By directly connecting the cable to one qubit in each node, we transfer quantum states between the nodes with a process fidelity of 0.911 +/- 0.008. We also prepare a three-qubit Greenberger-Horne-Zeilinger (GHZ) state in one node and deterministically transfer this state to the other node, with a transferred-state fidelity of 0.656 +/- 0.014. We further use this system to deterministically generate a globally distributed two-node, six-qubit GHZ state with a state fidelity of 0.722 +/- 0.021. In this work, the GHZ state fidelities are clearly above the threshold of 1/2 for genuine multipartite entanglement, showing that this architecture can be used to coherently link together multiple superconducting quantum processors, providing a modular approach for building large-scale quantum computers.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Megatux

The Megatux platform enables the emulation of large scale (multi-million node) distributed systems. In particular, it allows for the emulation of large-scale networks interconnecting a very large number of emulated computer systems. It does this by leveraging virtualization and associated technologies to allow hundreds of virtual computers to be hosted on a single moderately sized server or workstation. Virtualization technology provided by modern processors allows for multiple guest OSs to run at the same time, sharing the hardware resources. The Megatux platform can be deployed on a single PC, a small cluster of a few boxes or a large cluster of computers. With a modest cluster, the Megatux platform can emulate complex organizational networks. By using virtualization, we emulate the hardware, but run actual software enabling large scale without sacrificing fidelity.

Evensky, David↗

pnnl/eqc

EQC is a software framework for smart distributed quantum computer management and organization for variational quantum algorithms. EQC utilizes smart real time management of multiple independent quantum processors to maximize variational quantum algorithm throughput, for improved converged performance and improved rate of convergance.

Central, PNNL Developer↗

ADVANTG An Automated Variance Reduction Parameter Generator

The AutomateD VAriaNce reducTion Generator (ADVANTG) software automates the generation of variance reduction (VR) parameters for continuous-energy Monte Carlo simulations of fixed-source neutron, photon, and coupled neutron-photon transport problems using MCNP5 (X-5 Monte Carlo Team 2003). ADVANTG generates space- and energy dependent mesh-based weight-window bounds and biased source distributions from three dimensional (3-D) discrete ordinates (SN) calculations that are performed by the Denovo package (Evans et al. 2010). The deterministic calculations can be performed using multiple cores and/or processors (e.g., on multi-core desktop systems and clusters). The final variance reduction parameters are output in a format that can be used with unmodified versions of MCNP. The primary objective of the development of ADVANTG has been to reduce both the user effort and the computational time required to obtain accurate and precise tally estimates across a broad range of challenging transport application areas.

ADVANTG↗

Performance optimization, modeling and analysis of sparse matrix-matrix products on multi-core and many-core processors

Sparse matrix-matrix multiplication (SpGEMM) is a computational primitive that is widely used in areas ranging from traditional numerical applications to recent big data analysis and machine learning. Although many SpGEMM algorithms have been proposed, hardware specific optimizations for multi- and many-core processors are lacking and a detailed analysis of their performance under various use cases and matrices is not available. In this work, we firstly identify and mitigate multiple bottlenecks with memory management and thread scheduling on Intel Xeon Phi (Knights Landing or KNL). Specifically targeting many-core processors, we develop a hash-table-based algorithm and optimize a heap-based shared-memory SpGEMM algorithm. We examine their performance together with other publicly available codes. Different from the literature, our evaluation also includes use cases that are representative of real graph algorithms, such as multi-source breadth-first search or triangle counting. Our hash-table and heap-based algorithms are showing significant speedups from libraries in the majority of the cases while different algorithms dominate the other scenarios with different matrix size, sparsity, compression factor and operation type. We wrap up in-depth evaluation results and make a recipe to give the best SpGEMM algorithm for target scenario. We build the performance model for hash-table and heap-based algorithms, which supports the recipe. A critical finding is that hash-table-based SpGEMM gets a significant performance boost if the nonzeros are not required to be sorted within each row of the output matrix. Finally, we integrate our implementations into a large-scale protein clustering code named HipMCL, accelerating its SpGEMM kernel by up to 10X and achieving an overall performance boost for the whole HipMCL application by 2.6X.

97 MATHEMATICS AND COMPUTING↗

Analyzing the trade-off between multiple memory controllers and memory channels on multi-core processor performance

Increasing the core-count on current and future processors is posing critical challenges to the memory subsystem to efficiently handle concurrent memory requests. The current trend to cope with this challenge is to increase the number of memory channels available to the processor's memory controller. In this paper we investigate the effectiveness of this approach on the performance of parallel scientific applications. Specifically, we explore the trade-off between employing multiple memory channels per memory controller and the use of multiple memory controllers. Experiments conducted on two current state-of-the-art multicore processors, a 6-core AMD Istanbul and a 4-core Intel Nehalem-EP, for a wide range of production applications shows that there is a diminishing return when increasing the number of memory channels per memory controller. In addition, we show that this performance degradation can be efficiently addressed by increasing the ratio of memory controllers to channels while keeping the number of memory channels constant. Significant performance improvements can be achieved in this scheme, up to 28%, in the case of using two memory controllers with each with one channel compared with one controller with two memory channels.

97 MATHEMATICS AND COMPUTING↗

Approximation algorithms in combinatorial scientific computing

We survey recent work on approximation algorithms for computing degree-constrained subgraphs in graphs and their applications in combinatorial scientific computing. The problems we consider include maximization versions of cardinality matching, edge-weighted matching, vertex-weighted matching and edge-weighted $b$ -matching, and minimization versions of weighted edge cover and $b$ -edge cover. Exact algorithms for these problems are impractical for massive graphs with several millions of edges. For each problem we discuss theoretical foundations, the design of several linear or near-linear time approximation algorithms, their implementations on serial and parallel computers, and applications. Our focus is on practical algorithms that yield good performance on modern computer architectures with multiple threads and interconnected processors. We also include information about the software available for these problems.

Mathematics↗

Implementation of McMurchie–Davidson Algorithm for Gaussian AO Integrals Suited for SIMD Processors

We report an implementation of the McMurchie− Davidson evaluation scheme for 1- and 2-particle Gaussian AO integrals designed for processors with Single Instruction Multiple Data (SIMD) instruction sets. Like in our recent MD implementation for graphical processing units (GPUs) [Asadchev, A.; Valeev, E. F.. J. Chem. Phys. 2024, 160, 244109.], variable-sized batches of shellsets of integrals are evaluated at a time. By optimizing for the floating point instruction throughput rather than minimizing the number of operations, this approach achieves up to 50% of the theoretical hardware peak FP64 performance for many common SIMD-equipped platforms (AVX2, AVX512, NEON), which translates to speedups of up to 30 over the state-of-the-art one-shellset-at-a-time implementation of Obara−Saika-type schemes in Libint for a variety of primitive and contracted integrals. As with our previous work, we rely on the standard C++ programming language such as the std::simd standard library feature to be included in the 2026 ISO C++ standard without any explicit code generation to keep the code base small and portable. The implementation is part of the open source LibintX library freely available at https://github.com/ValeevGroup/libintx.

Basis sets↗

PIMS: Memristor-Based Processing-in-Memory-and-Storage.

Continued progress in computing has augmented the quest for higher performance with a new quest for higher energy efficiency. This has led to the re-emergence of Processing-In-Memory (PIM) ar- chitectures that offer higher density and performance with some boost in energy efficiency. Past PIM work either integrated a standard CPU with a conventional DRAM to improve the CPU- memory link, or used a bit-level processor with Single Instruction Multiple Data (SIMD) control, but neither matched the energy consumption of the memory to the computation. We originally proposed to develop a new architecture derived from PIM that more effectively addressed energy efficiency for high performance scientific, data analytics, and neuromorphic applications. We also originally planned to implement a von Neumann architecture with arithmetic/logic units (ALUs) that matched the power consumption of an advanced storage array to maximize energy efficiency. Implementing this architecture in storage was our original idea, since by augmenting storage (in- stead of memory), the system could address both in-memory computation and applications that accessed larger data sets directly from storage, hence Processing-in-Memory-and-Storage (PIMS). However, as our research matured, we discovered several things that changed our original direc- tion, the most important being that a PIM that implements a standard von Neumann-type archi- tecture results in significant energy efficiency improvement, but only about a O(10) performance improvement. In addition to this, the emergence of new memory technologies moved us to propos- ing a non-von Neumann architecture, called Superstrider, implemented not in storage, but in a new DRAM technology called High Bandwidth Memory (HBM). HBM is a stacked DRAM tech- nology that includes a logic layer where an architecture such as Superstrider could potentially be implemented.

97 MATHEMATICS AND COMPUTING↗

Performance Characteristics of the BlueField-2 SmartNIC

High-performance computing (HPC) researchers have long envisioned scenarios where application workflows could be improved through the use of programmable processing elements embedded in the network fabric. Recently, vendors have introduced programmable Smart Network Interface Cards (SmartNICs) that enable computations to be offloaded to the edge of the network. There is great interest in both the HPC and high-performance data analytics (HPDA) communities in understanding the roles these devices may play in the data paths of upcoming systems. This paper focuses on characterizing both the networking and computing aspects of NVIDIA’s new BlueField-2 SmartNIC when used in a 100Gb/s Ethernet environment. For the networking evaluation we conducted multiple transfer experiments between processors located at the host, the SmartNIC, and a remote host. These tests illuminate how much effort is required to saturate the network and help estimate the processing headroom available on the SmartNIC during transfers. For the computing evaluation we used the stress-ng benchmark to compare the BlueField-2 to other servers and place realistic bounds on the types of offload operations that are appropriate for the hardware. Our findings from this work indicate that while the BlueField-2 provides a flexible means of processing data at the network’s edge, great care must be taken to not overwhelm the hardware. While the host can easily saturate the network link, the SmartNIC’s embedded processors may not have enough computing resources to sustain more than half the expected bandwidth when using kernel-space packet processing. From a computational perspective, encryption operations, memory operations under contention, and on-card IPC operations on the SmartNIC perform significantly better than the general-purpose servers used for comparisons in our experiments. Therefore, applications that mainly focus on these operations may be good candidates for offloading to the SmartNIC.

97 MATHEMATICS AND COMPUTING↗

MPACT Development Activities at University of Michigan

This report documents the efforts of general development tasks in MPACT performed primarily at the University of Michigan during FY17. These tasks included: 1. Unstructured CMFD - The unstructured CMFD supports the RIA and BWR challenge problems. It allows MPACT to simulate mixed lattice cores (e.g. 8x8 next to 9x9). This design feature is common to operating BWRs and also the SPERT experiments being used to validate VERA-CS for RIA. 2. Higher order spatial sources for MOC - The higher order sources are for improving the overall performance of MPACT. Higher order spatial sources in MOC provide accurate solutions with potentially significantly fewer mesh regions, thus speeding up every calculation. This method is also particularly important for BWRs. 3. Improving fine source region visualization - The improvement of the visualization of the flat source region mesh in MPACT is to support users and communication of models and results from MPACT. VERAView provides visualization of pin-wise information, however MPACT resolves the spatial solution to intra pin level. The capability provided here will let users visualize the full model geometry and components of the solution at the highest spatial resolution available from MPACT. 4. Automated spatial decomposition - The automated spatial decomposition removes the burden from the user to know how to spatially decompose the model for MPACT; a task that has been rather tricky in the past. Instead the user may give some number of processors that is a multiple of the axial mesh and MPACT will determine a nearly optimal decomposition that gives better load balance than most decompositions used currently. 5. Optimal relaxation factors for feedback - The development of optimal relaxation factor is to improve the stability and robustness of problems with feedback. Presently the default feedback values are 0.5, which may or may not be optimal. Without relaxation the fixed point iteration that is currently used is known to be unstable for large problems. The optimal relaxation factors are derived by utilizing a previous Fourier analysis. From the original milestone there was a scope change where the work on the Intel MIC was deferred for work on the automated spatial decomposition.

22 GENERAL STUDIES OF NUCLEAR REACTORS↗

Reduction of background clutter in structured lighting systems

Methods for segmenting the reflected light of an illumination source having a characteristic wavelength from background illumination (i.e. clutter) in structured lighting systems can comprise pulsing the light source used to illuminate a scene, pulsing the light source synchronously with the opening of a shutter in an imaging device, estimating the contribution of background clutter by interpolation of images of the scene collected at multiple spectral bands not including the characteristic wavelength and subtracting the estimated background contribution from an image of the scene comprising the wavelength of the light source and, placing a polarizing filter between the imaging device and the scene, where the illumination source can be polarized in the same orientation as the polarizing filter. Apparatus for segmenting the light of an illumination source from background illumination can comprise an illuminator, an image receiver for receiving images of multiple spectral bands, a processor for calculations and interpolations, and a polarizing filter.

36 MATERIALS SCIENCE↗

In situ data acquisition and real-time analysis system

A testing system for evaluating the performance of an electrical/electronic UUT under dynamic operating conditions. The testing system includes a dynamic testing component (e.g., a centrifuge) for applying a stimulus to the UUT, and an iDAQ system configured to perform in situ data acquisition and real-time data analysis. The iDAQ system may also be subject to the stimulus. The iDAQ system includes a processor (e.g., an SoC) component, a power supply, a CR/I component, an IR component, and a single enclosure. The processor component may control the dynamic testing component, including varying in real-time the stimulus applied to the UUT. The processor component may include multiple input channels, and a high current/voltage subcomponent of the power supply may be configured to supply up to five hundred volts.

Benyhesan, Mohammad K.↗

A Novel Coarsening Method for Scalable and Efficient Mesh Generation

In this paper, we propose a novel mesh coarsening method called brick coarsening method. The proposed method can be used in conjunction with any graph partitioners and scales to very large meshes. This method reduces problem space by decomposing the original mesh into fixed-size blocks of nodes called bricks, layered in a similar way to conventional brick laying, and then assigning each node of the original mesh to appropriate brick. Our experiments indicate that the proposed method scales to very large meshes while allowing simple RCB partitioner to produce higher-quality partitions with significantly less edge cuts. Our results further indicate that the proposed brick-coarsening method allows more complicated partitioners like PT-Scotch to scale to very large problem size while still maintaining good partitioning performance with relatively good edge-cut metric. Graph partitioning is an important problem that has many scientific and engineering applications in such areas as VLSI design, scientific computing, and resource management. Given a graph G = (V,E), where V is the set of vertices and E is the set of edges, (k-way) graph partitioning problem is to partition the vertices of the graph (V) into k disjoint groups such that each group contains roughly equal number of vertices and the number of edges connecting vertices in different groups is minimized. Graph partitioning plays a key role in large scientific computing, especially in mesh-based computations, as it is used as a tool to minimize the volume of communication and to ensure well-balanced load across computing nodes. The impact of graph partitioning on the reduction of communication can be easily seen, for example, in different iterative methods to solve a sparse system of linear equation. Here, a graph partitioning technique is applied to the matrix, which is basically a graph in which each edge is a non-zero entry in the matrix, to allocate groups of vertices to processors in such a way that many of matrix-vector multiplication can be performed locally on each processor and hence to minimize communication. Furthermore, a good graph partitioning scheme ensures the equal amount of computation performed on each processor. Graph partitioning is a well known NP-complete problem, and thus the most commonly used graph partitioning algorithms employ some forms of heuristics. These algorithms vary in terms of their complexity, partition generation time, and the quality of partitions, and they tend to trade off these factors. A significant challenge we are currently facing at the Lawrence Livermore National Laboratory is how to partition very large meshes on massive-size distributed memory machines like IBM BlueGene/P, where scalability becomes a big issue. For example, we have found that the ParMetis, a very popular graph partitioning tool, can only scale to 16K processors. An ideal graph partitioning method on such an environment should be fast and scale to very large meshes, while producing high quality partitions. This is an extremely challenging task, as to scale to that level, the partitioning algorithm should be simple and be able to produce partitions that minimize inter-processor communications and balance the load imposed on the processors. Our goals in this work are two-fold: (1) To develop a new scalable graph partitioning method with good load balancing and communication reduction capability. (2) To study the performance of the proposed partitioning method on very large parallel machines using actual data sets and compare the performance to that of existing methods. The proposed method achieves the desired scalability by reducing the mesh size. For this, it coarsens an input mesh into a smaller size mesh by coalescing the vertices and edges of the original mesh into a set of mega-vertices and mega-edges. A new coarsening method called brick algorithm is developed in this research. In the brick algorithm, the zones in a given mesh are first grouped into fixed size blocks called bricks. These brick are then laid in a way similar to conventional brick laying technique, which reduces the number of neighboring blocks each block needs to communicate. Contributions of this research are as follows: (1) We have developed a novel method that scales to a really large problem size while producing high quality mesh partitions; (2) We measured the performance and scalability of the proposed method on a machine of massive size using a set of actual large complex data sets, where we have scaled to a mesh with 110 million zones using our method. To the best of our knowledge, this is the largest complex mesh that a partitioning method is successfully applied to; and (3) We have shown that proposed method can reduce the number of edge cuts by as much as 65%.

97 MATHEMATICS AND COMPUTING↗

Coupled‐cluster singles, doubles and perturbative triples with density fitting approximation for massively parallel heterogeneous platforms

Abstract A high‐performance implementation of the coupled‐cluster singles, doubles, and perturbative triples [CCSD(T)] is developed in the Massively Parallel Quantum Chemistry program. Novel features include: (1) reduced memory requirements via a density‐fitting (DF) CCSD implementation utilizing distributed lazy evaluation for tensors with more than two unoccupied indices and (2) the ability to utilize efficiently many‐core nodes (Intel Xeon Phi) and heterogeneous nodes with multiple NVIDIA GPUs on each node. All data that are greater than quadratic in the system size are distributed among processes. Excellent strong scaling is observed on distributed‐memory computers equipped with conventional CPUs, Intel Xeon Phi processors, and heterogeneous nodes with multiple NVIDIA GPUs Canonical CCSD(T) energies can be evaluated for systems containing 200 electrons and 1000 basis functions in a few days using a small size commodity cluster, with even larger computations possible on leadership‐class computing resources.

Peng, Chong↗

Developing an Automated Uncertainty Quantification Tool to Improve Watershed-Scale Predictions of Water and Nutrient Cycling

Managing the flow of water, nutrients, and contaminants in watersheds is vital to addressing pressing issues related to water scarcity, access to clean drinking water, energy production, resilience to natural and anthropogenic perturbations, and ecological restoration. Decisions about the management of watersheds critically depend on the accuracy with which the flow of water and chemicals through the watershed can be predicted by computer models. Prediction uncertainty can be reduced by matching the model to data, which are collected in the field at great expense. The contribution of watershed characterization data to reducing uncertainty of relevant model predictions can be evaluated in a so-called data-worth analysis, which provides transparent, quantitative metrics about a data set’s value for the support of relevant watershed management objectives. To achieve this goal, we developed a software package that implements the data-worth analysis approach for use with state-of-the-art watershed models. The purpose of the proposed data-worth analysis is to help decision-makers allocate resources for watershed characterization such that the uncertainty in model predictions can be significantly reduced, which leads to better, more effective management decisions. At the same time, watershed characterization costs can be reduced. The specific technical objectives of this SBIR/STTR Phase II project were to develop a framework and associated software toolsets that implement the uncertainty quantification and data-worth analysis approach for use with state-of-the-art watershed models. This goal was achieved by (A) developing a user-friendly, robust software package that is accessible to a wide audience, including watershed managers, policy-makers, and public stakeholders; (B) by demonstrating application of the prototype on several use cases that are representative of complex watershed management challenges spanning a range of scales and that consider different open-source, DOE-based codes and other modeling platforms; and (C) by gathering information about the needs and requirements from potential users to help guide future developments, ensuring that the final product will be commercially viable. The developed software consists of a graphical user interface that guides the user through a sequence of analysis steps, supported by toolsets that leverage state-of-the-art computational simulation-optimization capabilities. A prototype of the software runs on multiple platforms (PC, Mac, multi-processor Linux environment), is linked to diverse watershed simulators (e.g., ECOSYS, TOUGH2, TOUGHREACT, Amanzi-ATS), performs multiple analysis tasks (predictive simulations, sensitivity analysis, uncertainty analysis, automatic parameter estimation, and data-worth analysis, multicomponent geothermometry), and is readily extensible to include external simulators and analysis tools. The software is being commercialized and will be continually updated to address user needs.

58 GEOSCIENCES↗