OpenMP Model Research Articles

OpenMP-based parallel MLPG solver for analysis of heat conduction

PurposeIn the present work, we focus on developing an in-house parallel meshless local Petrov-Galerkin (MLPG) code for the analysis of heat conduction in two-dimensional and three-dimensional regular as well as complex geometries.Design/methodology/approachThe parallel MLPG code has been implemented using open multi-processing (OpenMP) application programming interface (API) on the shared memory multicore CPU architecture. Numerical simulations have been performed to find the critical regions of the serial code, and an OpenMP-based parallel MLPG code is developed, considering the critical regions of the sequential code.FindingsBased on performance parameters such as speed-up and parallel efficiency, the credibility of the parallelization procedure has been established. Maximum speed-up and parallel efficiency are 10.94 and 0.92 for regular three-dimensional geometry (343,000 nodes). Results demonstrate the suitability of parallelization for larger nodes as parallel efficiency and speed-up are more for the larger nodes.Originality/valueFew attempts have been made in parallel implementation of the MLPG method for solving large-scale industrial problems. Although the literature suggests that message-passing interface (MPI) based parallel MLPG codes have been developed, the OpenMP model has rarely been touched. This work is an attempt at the development of OpenMP-based parallel MLPG code for the very first time.

Predicting Software Defects in Hybrid MPI and OpenMP Parallel Programs Using Machine Learning

High-performance computing (HPC) and its supercomputers are essential for solving the most difficult issues in many scientific computing domains. The proliferation of computational resources utilized by HPC systems has resulted in an increase in the associated error rates. As such, modern HPC systems promote a hybrid programming style that integrates the message-passing interface (MPI) and open multi-processing (OpenMP). However, this integration often leads to complex defects, such as deadlocks and race conditions, that are challenging to detect and resolve. This paper presents a novel approach: using machine learning algorithms to predict defects in C++-based systems by employing hybrid MPI and OpenMP models. We focus on employing a balanced dataset to enhance prediction accuracy and reliability. Our study highlights the effectiveness of the support vector machine (SVM) classifier, enhanced with term frequency (TF) and recursive feature elimination (RFE) techniques, which demonstrates superior accuracy and performance in defect prediction when compared to other classifiers. This research contributes significantly to the field by providing a robust method for early defect detection in hybrid programming environments, thereby reducing development time, costs and improving the overall reliability of HPC systems.

Role-shifting threads: Increasing OpenMP malleability to address load imbalance at MPI and OpenMP

This paper presents the evolution of the free agent threads for OpenMP to the new role-shifting threads model and their integration with the Dynamic Load Balancing (DLB) library. We demonstrate how free agent threads can improve resource utilization in OpenMP applications with load imbalance in their nested parallel regions. We also demonstrate how DLB efficiently manages the malleability exposed by the role-shifting threads to address load imbalance issues. We use three real-world scientific applications, one of them to demonstrate that free agents alone can improve the OpenMP model without external tools, and two other MPI+OpenMP applications, one of them with a coupling case, to illustrate the potential of the free agent threads’ malleability with an external resource manager to increase the efficiency of the system. In addition, we demonstrate that the new implementation is more usable than the former one, letting the runtime system automatically make decisions that were made by the programmer previously. All software is released open-source.

Errors Classification and Static Detection Techniques for Dual-Programming Model (OpenMP and OpenACC)

Recently, incorporating more than one programming model into a system designed for high performance computing (HPC) has become a popular solution to implementing parallel systems. Since traditional programming languages, such as C, C++, and Fortran, do not support parallelism, many programmers add one or more programming models to achieve parallelism and accelerate computation efficiently. These models include Open Accelerators (OpenACC) and Open Multi-Processing (OpenMP), which have recently been used with various models, including Message Passing Interface (MPI) and Compute Unified Device Architecture (CUDA). Due to the difficulty of predicting the behavior of threads, runtime errors cannot be predicted. The compiler cannot identify runtime errors such as data races, race conditions, deadlocks, or livelocks. Many studies have been conducted on the development of testing tools to detect runtime errors when using programming models, such as the combinations of OpenACC with MPI models and OpenMP with MPI. Although more applications use OpenACC and OpenMP together, no testing tools have been developed to test these applications to date. This paper presents a testing tool for detecting runtime using a static testing technique. This tool can detect actual and potential runtime errors during the integration of the OpenACC and OpenMP models into systems developed in C++. This tool implement error dependency graphs, which are proposed in this paper. Additionally, a dependency graph of the errors is provided, along with a classification of runtime errors that result from combining the two programming models mentioned earlier.

Extensions of the two-phase double-point material point method to simulate the landslide-induced surge process

Landslide-induced surges, as a complex type of fluid-solid coupling problem, are widespread in mountain regions and cause disastrous consequences. In order to reproduce the entire process of landslide-induced surge, by taking full advantage of material point method (MPM) in simulating large deformation of soil and gravity-driven flow of water, the two-phase double-point material point method (TPDP-MPM) was extended: a new algorithm handling the boundary without setting boundary particles was proposed. The classical submarine block-induced surge test and sand column collapse experiment were numerically simulated by the MPM and verified the effectiveness of the program. After that, the multiphase coupling process of Lituya Bay landslide-induced surge was simulated and verified the reliability of the TPDP-MPM code, which have achieved parallel computing by making use of OpenMP model. By comparing the results of numerical simulation with the model experiment, the two-phase double-point material point method can well simulate the starting, propagation, run-up and reflux stage of landslide-induced surge. Moreover, the extension mechanism of landslide-surge disaster chain in both time and space is revealed from the perspective of energy. This study provides a reliable tool for the analysis and assessment of landslide-induced surge disasters.

An open-source solution to performance portability for Summit and Sierra supercomputers

Programming models that use a higher level of abstraction to express parallelism can target both CPUs and any attached devices, alleviating the maintainability and portability concerns facing today's heterogenous systems. This article describes the design, implementation, and delivery of a compliant OpenMP device offloading implementation for IBM-NVIDIA heterogeneous servers composing the Summit and Sierra supercomputers in the mainline open-source Clang/LLVM compiler and OpenMP runtime projects. From a performance perspective, reconciling the GPU programming model, best suited for massively parallel workloads, with the generality of the OpenMP model was a significant challenge. To achieve both high performance and full portability, we map high-level programming patterns to fine-tuned code generation schemes and customized runtimes that preserve the OpenMP semantics. In the compiler, we implement a low-overhead single-program multiple-data scheme that leverages the GPU native execution model and a fallback scheme to support the generality of OpenMP. Modular design enables the implementation to be extended with new schemes for frequently occurring patterns. Our implementation relies on key optimizations: sharing data among threads, leveraging unified memory, aggressive inlining of runtime calls, memory coalescing, and runtime simplification. We show that for commonly used patterns, performance on the Summit and Sierra GPUs matches that of hand-written native CUDA code.

The Method of Improving the Performance of Network Analysis Application for the Whole Power Grid

With the rapid development of UHV AC / DC hybrid power grid, it is required that the Network Analysis Application have the ability of unified analysis and high-performance computing. In this paper, the time-consuming analysis of each sub-function of Network Analysis Application is carried out to realize the performance bottleneck analysis of the whole process of network analysis. It is proposed that the performance of network analysis should be improved in three aspects: data input and output, topology analysis and core computing of Network Analysis Application. In the aspect of data input, the grid model data service is constructed to realize the combination of measurement reading and model reading, and the parallel model verification and boundary equivalence are completed; In the aspect of data output, it can be parallelized by table, block and equipment; In the aspect of topology analysis, the shared memory programming model OpenMP is adopted, and based on the fork/join parallel mode, its parallelization is realized; In the aspect of core computing, the existing parallel computing methods are summarized, and through the actual power grid simulation analysis, it puts forward the parallel computing mode applicable to different scale power grids and different applications. Finally, the effectiveness of this method is verified by comparing the optimized performance of state estimation.

A comparison of the shared-memory parallel programming models OpenMP, OpenACC and Kokkos in the context of implicit solvers for high-order FEM

We consider the application of three performance-portable programming models in the context of a high-order spectral element, implicit time-stepping solver for the Navier–Stokes equations. We aim to evaluate whether the use of these models allows code developers to deliver high-performance solvers for computational fluid dynamics simulations that are capable of effectively utilising both many-core CPU and GPU architectures. Using the core elliptic solver for the Navier–Stokes equations as a benchmarking guide, we evaluate the performance of these models on a range of unstructured meshes and give guidelines for the translation of existing codebases and their data structures to these models.

A Novel Approach for Efficient Training of Deep Neural Networks

<p style="text-indent: 1.27cm; margin-bottom: 0.35cm; line-height: 115%;" align="justify"><span style="font-family: Arial,serif;"><span style="font-size: small;"><em>Deep Neural Network training algorithms consumes long training time, especially when the number of hidden layers and nodes is large. Matrix multiplication is the key operation carried out at every node of each layer for several hundreds of thousands of times during the training of Deep Neural Network. Blocking is a well-proven optimization technique to improve the performance of matrix multiplication. Blocked Matrix multiplication algorithms can easily be parallelized to accelerate the performance further. This paper proposes a novel approach of implementing Parallel Blocked Matrix multiplication algorithms to reduce the long training time. The proposed approach was implemented using a parallel programming model OpenMP with collapse() clause for the multiplication of input and weight matrices of Backpropagation and Boltzmann Machine Algorithms for training Deep Neural Network and tested on multi-core processor system. Experimental results showed that the proposed approach achieved approximately two times speedup than classic algorithms.</em></span></span></p>

Portable Solution for Modeling Compressible Flows on All Existing Hybrid Supercomputers

A variant of a numerical algorithm for simulating viscous gasdynamic flows on unstructured hybrid grids and its software implementation for heterogeneous computations is described. The system of Navier–Stokes equations is approximated by the finite-volume method of an increased approximation order with the values of the variables being defined at the mass centers of the grid elements. The distributed software implementation of the numerical algorithm is adapted to running on hybrid computer systems of various architectures. Comparative implementations were created using the MPI, OpenMP, CUDA, and OpenCL software models permitting the use of multicore processors and various types of accelerators, including NVIDIA and AMD graphics processors, and Intel Xeon Phi multicore coprocessors. The data exchange between MPI processes and between processors and accelerators is carried out simultaneously with the execution of calculations (both in MPI + OpenMP mode and when using CUDA or OpenCL). The indicators of parallel efficiency and performance on systems with different types of computing devices are studied in detail. In the tests, up to 260 GPUs were successfully used.

The Research of the Parallel Computing Development from the Angle of Cloud Computing

Cloud computing is the development of parallel computing, distributed computing and grid computing. The development of cloud computing makes parallel computing come into people’s lives. Firstly, this paper expounds the concept of cloud computing and introduces two several traditional parallel programming model. Secondly, it analyzes and studies the principles, advantages and disadvantages of OpenMP, MPI and Map Reduce respectively. Finally, it takes MPI, OpenMP models compared to Map Reduce from the angle of cloud computing. The results of this paper are intended to provide a reference for the development of parallel computing.

Kernel optimization for short-range molecular dynamics

To optimize short-range force computations in Molecular Dynamics (MD) simulations, multi-threading and SIMD optimizations are presented in this paper. With respect to multi-threading optimization, a Partition-and-Separate-Calculation (PSC) method is designed to avoid write conflicts caused by using Newton’s third law. Serial bottlenecks are eliminated with no additional memory usage. The method is implemented by using the OpenMP model. Furthermore, the PSC method is employed on Intel Xeon Phi coprocessors in both native and offload models. We also evaluate the performance of the PSC method under different thread affinities on the MIC architecture. In the SIMD execution, we explain the performance influence in the PSC method, considering the “if-clause” of the cutoff radius check. The experiment results show that our PSC method is relatively more efficient compared to some traditional methods. In double precision, our 256-bit SIMD implementation is about 3 times faster than the scalar version.

基于多核DSP的OpenMp研究与实现

针对目前在实际应用场景中多核模型的复杂性，本文在分析了数据流、主从、OpenMp三大多核模型后，选取其中更易实现的OpenMp模型作为主要研究对象。本文研究了OpenMp模型的实现原理，对多核模型OpenMp在嵌套循环中的线程创建以及任务划分所消耗的时间进行了分析，然后研究了OpenMp模型的核数和线程数之间的关系，最终以TI公司的多核DSP TMS320C6678为核心处理器，使用典型的图像处理算法对在以上研究中得出的结论进行了验证。经测试，OpenMp模型在线程数等于核数时执行时间最短。 Aiming at the complexity of multicore model in practical application, the three major models of data flow, master-slave and OpenMp are analyzed in this paper. Because the OpenMp model is easier to be implemented, it is taken as the major research object. Firstly, the implementation principle of OpenMp model is studied; the time consumed by the thread creation and task parti-tioning in the nested loop of the OpenMp model is analyzed in this paper. Then, the relationship between the core number and the number of threads in the OpenMp model is also studied. Finally, taking TMS320C6678 multi-core DSP of TI Company as the core processor, a simple image processing algorithm is used to verify the conclusions drawn from the above research. The results of test have proved OpenMp model when the number of threads is equal to the core number shortest execution time.

Ariadne — Directive-based parallelism extraction from recursive functions

In this paper we present Ariadne, a compiler that extracts parallelism from recursive function calls. Ariadne takes as input C code enhanced with directives for recursive functions and automatically produces code for multi-core architectures. It produces code for the POSIX standard, the OpenMP model and the Cilk programming language, which run on a wide variety of computing systems. Ariadne also produces code for SL, a programming language proposed for the SVP processor and model. This is of special interest, since we can map certain function calls onto SVP, which contain inherent parallelism that cannot efficiently be expressed in other programming models. Ariadne is the only compiler that extracts parallelism from various forms of recursive functions using directives. It is also the only compiler that handles all forms of reduction operations for addition, subtraction, multiplication and division. The experimental results are very promising showing significant speedups in all benchmarks.

Parallel Implementation of Singular Value Decomposition (SVD) in Image Compression using Open Mp and Sparse Matrix Representation

The advent of Multi-core processors has offered powerful processing capabilities and provided new avenues for parallel processing. As the traditional methods of computation are inherently based on single core of execution, they are not capable of taking the advantage of high computational power offered by multi core processors, even if available. Singular Value Decomposition (SVD) has been proven well for many image processing techniques such as image compression, quality measure and in watermarking. SVD is a highly compute intensive algorithm that applies numerous matrix operations to an image such as transpose, inverse, multiplication of high orders to form a compressed image. However accelerating the SVD routines to exploit the underlying hardware poses a significant challenge to the programmers. This paper deals with improving the speedup of SVD algorithm used in image compression technique. This is achieved by identifying the areas where data parallelism can be applied and parallel programming model OpenMp is used to run the parallel components on multiple processors. This results in faster execution of the algorithm and hence reduces the execution time when applied to image compression for large images. This paper also addresses the space overhead of storing the matrices of SVD technique, by adapting efficient sparse matrix storage format such as Compressed Row Storage (CSR). Experimental results show that the speedup achieved through OpenMp is around 1.15 and better compression ratio with sparse matrix storage format.

The Parallel Solving and Program Implementation Method of Large-Scale Power Flow Correction Equations

This paper present a parallel solving method of large-scale power flow correction equations and the program implementation based on the shared memory programming model OpenMP, which Improve the utilization rate the computer CPU resources and the computing speed. The method uses factor tables for solving the power flow equations. Combined the sparsity of correction equations coefficient matrix, the rows which can be do normalized simultaneously are grouped for doing the normalization and elimination operations in parallel. Meanwhile, use the row grouping information which is obtained during the factor table generation process to do parallel previous generation computing. Finally, the simulation for the actual power grid verified the validity and rationality of the method.

Programming parallel embedded and consumer applications in OpenMP superscalar

In this paper, we evaluate the performance and usability of the parallel programming model OpenMP Superscalar (OmpSs), apply it to 10 different benchmarks and compare its performance with corresponding POSIX threads implementations.

A real-time tracking method for the rectangular magnet based on parallel Levenberg-Marquardt algorithm

magnetic flux intensity of the rectangular magnet is the function of its location and orientation. With the measured magnetic flux intensities at two points around the rectangular magnet by 3-axis magnetic sensors, a non-linear equation group is established, and the location and orientation of the rectangular magnet can be determined by seeking the solution of the equation group by using general Levenberg-Marquardt (L-M) algorithm. However, the time occupied by general L-M algorithm is too long to realize real-time tracking for the rectangular magnet. We propose an improved method which computes Jacobin Matrix in L-M algorithm in parallel. Making use of the Open Multi-Processing (OpenMP) model, the parallel operation is implemented on the dual-core personal computer. As a result, the real-time tracking for the rectangular magnet is realized, and the tracking curve is smooth.

Designing parallel loop self-scheduling schemes using the hybrid MPI and OpenMP programming model for multi-core grid systems

Loop scheduling on parallel and distributed systems has been thoroughly investigated in the past. However, none of these studies considered the multi-core ar- chitecture feature for emerging grid systems. Although there have been many studies proposed to employ the hybrid MPI and OpenMP programming model to exploit dif- ferent levels of parallelism for a distributed system with multi-core computers, none of them were aimed at parallel loop self-scheduling. Therefore, this paper investigates how to employ the hybrid MPI and OpenMP model to design a parallel loop self- scheduling scheme adapted to the multi-core architecture for emerging grid systems. Three different featured applications are implemented and evaluated to demonstrate the effectiveness of the proposed scheduling approach. The experimental results show that the proposed approach outperforms the previous work for the three applications and the speedups range from 1.13 to 1.75.

Performance evaluation of programming models for SMP‐based clusters

Recently, computing clusters based on shared‐memory multiprocessors (SMP's) is becoming popular for high performance computing (HPC) applications. With the recent prevalence of CPU's, which are small‐scale SMP's themselves, multi‐core CPU's SMP clusters will become increasingly popular in the near future. SMP clusters have characteristics of both SMP's and MPP's. Therefore, developing parallel programs which can efficiently exploits characteristics of both SMP and MPP in SMP clusters is a challenging task. Standard parallel programming models such as MPI, OpenMP, or Hybrid (a combination of the two former models) are commonly used for SMP clusters. Depending on the characteristics of applications, however, some programming models are better than others. To identify and select a suitable programming model for an application on SMP clusters needs a quantity of analysis of the application behavior and its performance. In this paper, we conduct experimental studies to evaluate the benefits and limits of MPI and OpenMP on three SMP‐based systems using standard HPC applications parallelized using MPI, OpenMP, and Hybrid model. The performance results and final analysis may lead to an optimal programming model for the applications.