Parallel Computer Architectures Research Articles

Bayesian uncertainty quantification and propagation for discrete element simulations of granular materials

Predictions in the behavior of granular materials using Discrete Element Methods (DEM) hinge on the employed interaction potentials. Here we introduce a data driven, Bayesian framework to quantify DEM predictions. Our approach relies on experimentally measured coefficients of restitution for single steel particle–wall collisions. The calibration data entail both tangential and normal coefficients of restitution, for varying impact angles and speeds of the bouncing particle. The parametric uncertainty in multiple Force–Displacement models is estimated using an enhanced Transitional Markov Chain Monte Carlo implemented efficiently on parallel computer architectures. In turn, the parametric model uncertainties are propagated to predict Quantities of Interest (QoI) for two testbed applications: silo discharge and vibration induced mass-segregation. This work demonstrates that the classical way of calibrating DEM potentials, through parameter optimization, is insufficient and it fails to provide robust predictions. The present Bayesian framework provides robust predictions for the behavior of granular materials using DEM simulations. Most importantly the results demonstrate the importance of including parametric and modeling uncertainties in the potentials employed in Discrete Element Methods.

Parallel Compression and Decompression of DNA Sequence Reads in FASTQ Format

Large volumes of short reads of genomic data are generated by high-throughput sequencing instruments. The FASTQ format is widely accepted as the input format of genomic reads and has presented challenges in data storage, management, and transfer. The performance of this type of serial algorithms such as G-SQZ and DSRC is limited by the single processor and the memory in a single computer. Utilizing data parallelism, the circular dual queues of buffers, memory mapping integrated with superblocks, pipeline parallelism with multi threads, and so on, we present the parallel compression and decompression methods for DNA sequence reads in FASTQ format based on the parallel computer architectures of the cluster and the SMP. Experimental results for the parallel DSRC algorithm clearly show the efficiency of using the powerful computing resources from multi computing nodes and multi cores of each node. The speedups vary from 46 to 62 for parallel compression and vary from 40 to 58 for parallel decompression by using 10 nodes of the cluster in Tianhe-1A super computer. Test results on the SMP machine are also pleasant. The methods could be applied to any serial compressing algorithms of DNA sequence reads in FASTQ format only if they have the traits of index and superblocks.

Data Processing and Applied Technology in the Algorithm and Cracking of NTLM-Hash - A Highly Proficient Parallel Computing Architecture

Compared with CPU, GPU has a stronger parallel architecture and with specific API, the powerful potential contained in GPU can be released to become GPGPU. High computational capability of GPGPU can be used in scientific computation and data processing. This paper focuses on the research of the high frequency collision crack for normal Windows NTLM-Hash by resorting to CUDA API. The result shows that the computation capability of low-end graphic cards could be almost 100 times that of the normal dual-core CPUs, which could greatly lift the efficiency of encryption cracking.

Hybrid parallel computing architecture for multiview phase shifting

Abstract. The multiview phase-shifting method shows its powerful capability in achieving high resolution three-dimensional (3-D) shape measurement. Unfortunately, this ability results in very high computation costs and3-D computations have to be processed offline. To realize real-time 3-D shape measurement, a hybrid parallelcomputing architecture is proposed for multiview phase shifting. In this architecture, the central processingunit can co-operate with the graphic processing unit (GPU) to achieve hybrid parallel computing. The highcomputation cost procedures, including lens distortion rectification, phase computation, correspondence,and 3-D reconstruction, are implemented in GPU, and a three-layer kernel function model is designed to simul-taneously realize coarse-grained and fine-grained paralleling computing. Experimental results verify that thedeveloped system can perform 50 fps (frame per second) real-time 3-D measurement with 260 K 3-D pointsper frame. A speedup of up to 180 times is obtained for the performance of the proposed technique using aNVIDIA GT560Ti graphics card rather than a sequential C in a 3.4 GHZ Inter Core i7 3770.

LumiNOC: A Power-Efficient, High-Performance, Photonic Network-on-Chip

To meet energy-efficient performance demands, the computing industry has moved to parallel computer architectures, such as chip multiprocessors (CMPs), internally interconnected via networks-on-chip (NoC) to meet growing communication needs. Achieving scaling performance as core counts increase to the hundreds in future CMPs, however, will require high performance, yet energy-efficient interconnects. Silicon nanophotonics is a promising replacement for electronic on-chip interconnect due to its high bandwidth and low latency, however, prior techniques have required high static power for the laser and ring thermal tuning. We propose a novel nano-photonic NoC (PNoC) architecture, LumiNOC, optimized for high performance and power-efficiency. This paper makes three primary contributions: a novel, nanophotonic architecture which partitions the network into subnets for better efficiency; a purely photonic, in-band, distributed arbitration scheme; and a channel sharing arrangement utilizing the same waveguides and wavelengths for arbitration as data transmission. In a 64-node NoC under synthetic traffic, LumiNOC enjoys 50% lower latency at low loads and ${\sim}{40\%}$ higher throughput per Watt on synthetic traffic, versus other reported PNoCs. LumiNOC reduces latencies ${\sim}{40\%}$ versus an electrical 2-D mesh NoCs on the PARSEC shared-memory, multithreaded benchmark suite.

A scalable parallel genetic algorithm for the Generalized Assignment Problem

• A scalable parallel genetic algorithm (PGA) is developed for the NP-hard GAP problem. • High scalability is achieved through a novel asynchronous migration strategy. • Our algorithmic analysis resolves the buffer overflow issue in asynchronous PGAs. • Strong and weak scaling tests demonstrate the superiority of our PGA approach. • Our approach scales to 16,384 processors with super-linear speedups observed. Known as an effective heuristic for finding optimal or near-optimal solutions to difficult optimization problems, a genetic algorithm (GA) is inherently parallel for exploiting high performance and parallel computing resources for randomized iterative evolutionary computation. It remains to be a significant challenge, however, to devise parallel genetic algorithms (PGAs) that can scale to massively parallel computer architecture (also known as the mainstream supercomputer architecture) primarily because: (1) a common PGA design adopts synchronized migration, which becomes increasingly costly as more processor cores are involved in global synchronization; and (2) asynchronous PGA design and associated performance evaluation are intricate due to the fact that PGA is a type of stochastic algorithm and the amount of computation work needed to solve a problem is not simply dependent on the problem size. To address the challenge, this paper describes a scalable coarse-grained PGA–PGAP, for a well-known NP -hard optimization problem: Generalized Assignment Problem (GAP). Specifically, an asynchronous migration strategy is developed to enable efficient deme interactions and significantly improve the overlapping of computation and communication. Buffer overflow and its relationship with migration parameters were investigated to resolve the issues of observed message buffer overflow and the loss of good solutions obtained from migration. Two algorithmic conditions were then established to detect these issues caused by communication delays and improper configuration of migration parameters and, thus, guide the dynamic tuning of PGA parameters to detect and avoid these issues. A set of computational experiments is designed to evaluate the scalability and numerical performance of PGAP. These experiments were conducted for large GAP instances on multiple supercomputers as part of the National Science Foundation Extreme Science and Engineering Discovery Environment (XSEDE). Results showed that, PGAP exhibited desirable scalability by achieving low communication cost when using up to 16,384 processor cores. Near-linear and super-linear speedups on large GAP instances were obtained in strong scaling tests. Desirable scalability to both population size and the number of processor cores were observed in weak scaling tests. The design strategies applied in PGAP are applicable to general asynchronous PGA development.

Multi-Agent Data Fusion System Based on Multiple–FPGA Parallel Computing Architecture

Multi-Agent Data Fusion System Based on Multiple–FPGA Parallel Computing Architecture

Complex Modeling of Matrix Parallel Algorithms

Parallel principles are the most effective way how to increase performance in parallel computing (parallel computers and algorithms too). In this sense the paper is devoted to a complex performance evaluation of matrix parallel algorithms (MPA). At first the paper describes the typical matrix parallel algorithms and then it summarizes common properties of them to complex performance modeling of MPA. To complex performance analysis we are able to take into account all overheads influence performance of parallel algorithms (parallel computer architecture, parallel computation, communication etc.). To be le to analyze MPA in their abstract form we have defined needed decomposition models of MPA. For these decomposition strategies we derived analytical relation for defined complex performance criterions including isoefficiency functions, which allow us to predict performance although for hypothetical parallel computer. In its experimental part the paper considers the achieved results using defined complex performance criterions including issoefficiency function for performance prediction also for hypothetical future parallel computers. Such idea of common abstract analysis could be very useful in deriving complex performance criterions for groups of other similar parallel algorithms (PA) as for example numerical integration PA, optimization PA etc.

A High Performance Optoelectronic Machine for Automated Arabic-English Parallel Corpus Creation and for Text Mining Processing

 Abstract—In this paper a parallel optoelectronic computer architecture is proposed for large-scale parallel corpus, full text search and text mining applications while achieving high speed and high performance and utilizing the parallel processing nature of optics. An optical technique is presented that aims to automatically construct an Arabic-English corpus through web mining. In addition, it is shown how Arabic morphology and Arabic synonymous can be used to extract the required knowledge from Arabic texts based on converting the Arabic text into a semi-structured database. The proposed system is modeled and implemented using a two-dimensional space optics symbolic correlator to perform the comparison operations of the expert system. The proposed optical engine is able to perform 3D processing for inferring conclusions to queries by representing page contents in a 2D space using mapped-template logic. A hardware realization for these systems based on commercially available components is discussed. The performance analysis of the proposed system shows its remarkably improved processing performance with respect to processing speeds and both recall and precision measures.

Null steering in irregularly spaced sparse antenna arrays using aperture distributed subarrays and hybrid optimiser

Sparse antenna arrays with irregularly spaced elements are increasingly being used in large array applications such as radio astronomy and radar. These antenna arrays may consist of hundreds of elements and therefore optimisation of these arrays becomes computationally expensive. This study proposes a novel solution to the problem of null synthesis in such arrays, where nulls are formed in prescribed locations for two-dimensional irregular sparse antenna arrays. The technique is based on a subarray and hybrid synthesis solution, where each subarray spans the entire antenna aperture. Single and multiple null formations with array gain maximisation are considered. The technique's effectiveness is demonstrated using evolutionary optimisation algorithms; however, the basic technique can be used with other optimisation methods. The simulation results presented compare the null synthesis performance with and without subarraying and it has shown that the use of subarray results in improved convergence compared with a direct full array synthesis. Furthermore the technique is appropriate for ready application to parallel computer architectures.

Modeling of Single Computing Nodes of Parallel Computers

The paper describes analytical modeling for single computing nodes of parallel computers. At first the paper describes very shortly the developing steps of parallel computer architectures and then he summarized the basic concepts for performance evaluation. To illustrate theoretical evaluation concepts the paper considers in its experimental part the achieved results on concrete analyzed examples and their comparison. The suggested analytical models consider for single computing node based on processor or core and SMP modeling of own computer node´s activities and node´s communication channels of performed data communications within computing node queuing theory systems M/D/m or M/D/. In case of using SMP parallel system as node computer the suggested models consider for own node’s activities M/M/m or M/D/m queuing theory systems. Although we are able to use other more complicated queuing theory systems we prefer modeling with mentioned models because achieved results for these models we can use in decomposed modeling of coupled computing nodes as network of workstations (NOW) or network of massive NOW modules (Grid). The achieved results of the developed analytical models we have compared with the results of tested computing nodes with other alternative evaluation method based on suitable benchmarks to verify developed analytical models. The developed analytical models could be used under various ranges of input analytical parameters, which influence the architecture of analyzed computing nodes which are interested for the praxis.

A Cloud Computing Service Architecture of a Parallel Algorithm Oriented to Scientific Computing with CUDA and Monte Carlo

Abstract The GPGPU (General Purpose Graphics Processing Units) have become a whole new area for research due to the fast development of GPU hardware and programming tools, such as CUDA (Compute Unified Device Architecture). Here we have made a research on CUDA and its applications in the field of scientific computation, as organ electronics. We propose one solution of the parallel computation in global optimization of the physical characteristics in organ electronics with Monte Carlo, and one cloud service architecture for parallel computation of organ electronics was designed. Finally, one case of computing the organ molecule moving orbit was implemented based on the above solution and architecture, and has got a good display by the cloud service.

Beyond standard molecular dynamics: investigating the molecular mechanisms of G protein-coupled receptors with enhanced molecular dynamics methods.

The majority of biological processes mediated by G Protein-Coupled Receptors (GPCRs) take place on timescales that are not conveniently accessible to standard molecular dynamics (MD) approaches, notwithstanding the current availability of specialized parallel computer architectures, and efficient simulation algorithms. Enhanced MD-based methods have started to assume an important role in the study of the rugged energy landscape of GPCRs by providing mechanistic details of complex receptor processes such as ligand recognition, activation, and oligomerization. We provide here an overview of these methods in their most recent application to the field.

Parallel Algorithm for DC Resistivity Method Based on CUDA-Enabled GPU

During the electrical prospecting, the three-dimensional forward problem has been the hot topic in the research of DC electrical method. When the forward computation results are solved through the finite element method and the finite difference method, a large-scale sparse linear equation set should be obtained, moreover the computation in the solution of large-scale linear algebraic equation sets are very heavy. If we adopt serial computation, the computing efficiency is very low, which greatly affects the application efficiency. With the increasing maturation of the parallel computer architecture, the Graphic Processing Unit (GPU) parallel computing modeling can apply in this field to enable the efficiency of the three-dimensional forward modeling to be significantly improved.

A Cellwise Block-Gauss-Seidel Iterative Method for Multigroup SN Transport on a Hybrid Parallel Computer Architecture

A Fourier analysis is conducted in two-dimensional (2-D) geometry for the discrete ordinates (SN) approximation of the neutron transport problem solved with Richardson iteration (source iteration) using the cellwise block-Jacobi (bJ) and block-Gauss-Seidel (bGS) algorithms. The results of the Fourier analysis show that convergence of bJ can degrade, leading to a spectral radius ρ equal to 1, in problems containing optically thin cells. For problems containing cells that are optically thick, instead, ρ tends to 0. Hence, in the optically-thick-cell regime, bJ is rapidly convergent even for scattering-dominated problems. Similar conclusions hold for bGS, except bGS approaches the asymptotic, thick-cell regime at convergence rates higher than bJ. Hence, we have implemented the bGS algorithm on the Roadrunner hybrid, parallel computer architecture. A compute node of this massively parallel machine comprises AMD Opteron cores that are linked to a Cell Broadband Engine (Cell/B.E.). LAPACK routines have been ported to the Cell/B.E. in order to make use of its parallel synergistic processing elements (SPEs). The bGS algorithm is based on the LU factorization and solution of a linear system that couples the fluxes for all SN angles and energy groups on a mesh cell. For every cell of a mesh that has been parallel decomposed on the higher-level Opteron processors, a linear system is transferred to the Cell/B.E. and the parallel LAPACK routines are used to compute a solution, which is then transferred back to the Opteron, where the rest of the SN transport computations take place. Compared to standard parallel machines, a one-hundred-fold speedup of the bGS was observed on Roadrunner. Numerical experiments with strong and weak parallel scaling demonstrate that the bGS method is viable and compares favorably to full parallel transport sweeps (FPS) on 2-D unstructured meshes when it is applied to optically thick, multimaterial problems. Specifically, the strong parallel efficiency of accelerated bGS on Roadrunner can achieve 73% at 512 processors, compared with 32 processors, while efficiency is 34% for the (Opteron-only) implementation of FPS. The weak parallel efficiency of bGS is 58% while it reaches 10% for FPS. As expected, however, bGS is not as efficient as FPS in optically thin problems.

Simulation of single-phase mixing in fuel rod bundles, using an immersed boundary method

Over recent decades, the study of turbulent flow structures and the mixing behavior in fuel rod bundles has been an active research topic. In this paper, the (isothermal) turbulent mixing in rod bundles is simulated for two different cases using a finite-volume/immersed boundary method with a staggered velocity discretization. The solution procedure is based on an efficient multigrid algorithm, and simulations are performed on state-of-the-art massively parallel computer architectures. As the first case, the cross-flow through a staggered tube bundle under fully turbulent conditions is computed using large eddy simulation. The computed velocities were compared with the experimental data from the literature, and the agreement between the experimental and the computational results was found to be rather good. As a second case, we considered an experiment performed in the SUBFLOW test facility. The SUBFLOW facility models a 4 × 4 vertical rod bundle and uses water in atmospheric conditions as the working fluid. Experiments are based on the wire-mesh sensor technique, which measures the conductivity of the working fluid, and report the distribution of the tracer inserted at various heights of the bundle. This case is even more challenging for large eddy simulations, since it does not feature large structures present in the flow crossing the bundles. However, large eddy simulations of the SUBFLOW facility reproduced the concentration of the tracer liquid and its fluctuations well. The good agreement obtained for the two cases of flows in rod bundles serves as a motivation to focus future research on the modeling of bubbly flows in the same arrangements.

Perspective of Parallel Simulation in Mechatronic Systems - Metal Cutting

The method of simulation design at the computer systems is discussed. In the system identification theory of systems, modelling a program implementation is presented. The simulation of systems depends on definition of their mathematical models. The simulation method is rooted in the fact it is an experimental method comprising aspect like model calibration and data collection. These models are presented in a matrix form in a very good way. A matrix form is often employed to define tasks in the field of science and engineering.. The result of simulation of mathematical models of mechatronic devices, control and communication technology, control, climate changes and etc. can be created using parallel computer architecture. Advantages for users are presented by practical example of cutting process associated with the simulation of the observer.

Interactive scheduling for mobile multimedia service in M2M environment

Computational load of motion estimation in advanced video coding (AVC) standard is significantly high and its more true for HDTV sequences. In this paper, video processing algorithm is mapped onto a learning method to improve machine to machine (M2M) architecture, namely, the parallel reconfigurable computing (PRC) architecture, which consists of multiple units, First, we construct a directed acyclic graph (DAG) to represent the video coding algorithms comprising motion estimation. In the future trillions of devices are connected (M2M) together to provide services and that time power management would be a challenge. Computation aware scheme for different machine is reduced by dynamically scheduling usage of multi-core processing environment for video sequence depending up complexity of the video. And different video coding algorithm is selected depending upon the nature of the video. Simulation results show the effectiveness of the proposed method.

Toward GPU accelerated topology optimization on unstructured meshes

The present work investigates the feasibility of finite element methods and topology optimization for unstructured meshes in massively parallel computer architectures, more specifically on Graphics Processing Units or GPUs. Challenges in the parallel implementation, like the parallel assembly race condition, are discussed and solved with simple algorithms, in this case greedy graph coloring. The parallel implementation for every step involved in the topology optimization process is benchmarked and compared against an equivalent sequential implementation. The ultimate goal of this work is to speed up the topology optimization process by means of parallel computing using off-the-shelf hardware. Examples are compared with both a standard sequential version of the implementation and a massively parallel version to better illustrate the advantages and disadvantages of this approach.

An algorithm for massively parallel dislocation dynamics simulations of small scale plasticity

Accurate modeling of dislocation motion in bounded bodies is essential for the goal of obtaining desired properties, for example electronic or optical, of many microelectronic devices. At present, we lack high fidelity computer codes for such modeling that efficiently utilize modern parallel computer architectures. In contrast, many dislocation simulation codes are available for periodic or infinite bodies. In principle, these codes can be extended to allow for dislocation modeling in finite bodies. However, such extension may involve an additional solver to be employed, coupled with a dislocation simulation code. We present an algorithm for development of parallel dislocation simulation capability for bounded bodies based on such coupling. Subsequently, we analyze the performance of the algorithm for a demanding dislocation dynamics model problem.