Graphics Processing Unit Systems Research Articles

Accelerating parallel particle swarm optimization via GPU

Particle swarm optimization (PSO) is a population-based stochastic and derivative-free method that has been used to solve various optimization problems due to its simplicity and efficiency. While solving high-dimensional or complicated problems, PSO requires a large number of particles to explore the problem domains and consequently introduces high computational costs. In this paper, we focus on the acceleration of PSO for solving box-constrained, load-balanced optimization problems by parallelization on a graphics processing unit (GPU). We propose a GPU-accelerated PSO (GPSO) algorithm by using a thread pool model and implement GPSO on a GPU. Numerical results show that the GPU architecture fits the PSO framework well by reducing computational timing, achieving high parallel efficiency and finding better optimal solutions by using a large number of particles. For example, while solving the 100-dimensional test problems with 65,536 (16×212) particles, GPSO has achieved up to 280X and 83X speedups on a NVIDIA Tesla C1060 1.30 GHz GPU relative to an Intel Xeon-X5450 3.00 GHz central processing unit running in single- and quad-core mode, respectively. GPSO provides a promising method for tackling high-dimensional and difficult optimization problems using a low-cost and many-core GPU system.

Does Parametric fMRI Analysis with SPM Yield Valid Results? – An Empirical Study of 1484 Rest Datasets

Does Parametric fMRI Analysis with SPM Yield Valid Results? – An Empirical Study of 1484 Rest Datasets

Performance potential for simulating spin models on GPU

Graphics processing units (GPUs) are recently being used to an increasing degree for general computational purposes. This development is motivated by their theoretical peak performance, which significantly exceeds that of broadly available CPUs. For practical purposes, however, it is far from clear how much of this theoretical performance can be realized in actual scientific applications. As is discussed here for the case of studying classical spin models of statistical mechanics by Monte Carlo simulations, only an explicit tailoring of the involved algorithms to the specific architecture under consideration allows to harvest the computational power of GPU systems. A number of examples, ranging from Metropolis simulations of ferromagnetic Ising models, over continuous Heisenberg and disordered spin-glass systems to parallel-tempering simulations are discussed. Significant speed-ups by factors of up to 1000 compared to serial CPU code as well as previous GPU implementations are observed.

Guest Editor's Introduction: Special Section on Challenges and Solutions in Multicore and Many‐Core Computing

Guest Editor's Introduction: Special Section on Challenges and Solutions in Multicore and Many‐Core Computing

Fine-grain parallelism using multi-core, Cell/BE, and GPU Systems

Currently, we are facing a situation where applications exhibit increasing computational demands and where a large variety of parallel processor systems are available. In this paper we focus on exploiting fine-grain parallelism for three applications with distinct characteristics: a Bioinformatics application (MrBayes), a Molecular Dynamics application (NAMD), and a database application (TPC-H). We assess, side-by-side, the performance of the three applications on general-purpose multi-core processors, the Cell Broadband Engine (Cell/BE), and Graphics Processing Units (GPU). Our results indicate that application performance depends on the characteristics of the parallel architectures and on the computational requirements of the core functions of the respective applications. For MrBayes the best overall performance is achieved on general-purpose multi-core processors, for NAMD on the Cell/BE, and for TPC-H on GPUs.

Solving finite difference linear systems on GPUs: CUDA based Parallel Explicit Preconditioned Biconjugate Conjugate Gradient type Methods

During the last decades, explicit approximate inverse preconditioning methods have been used for efficiently solving sparse linear systems on multiprocessor systems. The effectiveness of explicit approximate inverse preconditioning schemes relies on the use of efficient preconditioners that are close approximants to the coefficient matrix and are fast to compute in parallel. A new parallel computational technique is proposed for the parallelization of the explicit preconditioned conjugate gradient type method on a Graphics Processing Unit (GPU). The proposed parallel methods have been implemented using Compute Unified Device Architecture (CUDA) developed by NVIDIA. The inherently parallel operations between vectors and matrices involved in the explicit preconditioned biconjugate conjugate gradient type schemes exhibit significant amounts of loop-level parallelism because of the matrix–vector and the vector–vector products that can lead to high performance gain on the GPU systems, specifically designed for such computations. Finally, numerical results for the performance of the explicit preconditioned biconjugate conjugate gradient type method for solving characteristic two dimensional boundary value problems, using the finite difference method, on a massive multiprocessor interface on a GPU are presented. The CUDA implementation issues of the proposed method are also discussed.

Implementations of a Parallel Algorithm for Computing Euclidean Distance Map in Multicore Processors and GPUs

Given a 2-D binary image of size n×n, Euclidean Distance Map (EDM) is a 2-D array of the same size such that each element is storing the Euclidean distance to the nearest black pixel. It is known that a sequential algorithm can compute the EDM in O(n2) and thus this algorithm is optimal. Also, work-time optimal parallel algorithms for shared memory model have been presented. However, the presented parallel algorithms are too complicated to implement in existing shared memory parallel machines. The main contribution of this paper is to develop a simple parallel algorithm for the EDM and implement it in two different parallel platforms: multicore processors and Graphics Processing Units (GPUs). We have implemented our parallel algorithm in a Linux server with four Intel hexad-core processors (Intel Xeon X7460 2.66GHz). We have also implemented it in the following two modern GPU systems, Tesla C1060 and GTX 480, respectively. The experimental results have shown that, for an input binary image with size of 9216×9216, our implementation in the multicore system achieves a speedup factor of 18 over the performance of a sequential algorithm using a single processor in the same system. Meanwhile, for the same input binary image, our implementation on the GPU achieves a speedup factor of 26 over the sequential algorithm implementation.

Higher-order CFD and interface tracking methods on highly-Parallel MPI and GPU systems

A computational investigation of the effects on parallel performance of higher-order accurate schemes was carried out on two different computational systems: a traditional CPU based MPI cluster and a system of four Graphics Processing Units (GPUs) controlled by a single quad-core CPU. The investigation was based on the solution of the level set equations for interface tracking using a High-Order Upstream Central (HOUC) scheme. Different variants of the HOUC scheme were employed together with a 3rd-order TVD Runge–Kutta time integration. An increase in performance of two orders of magnitude was seen when comparing a single CPU core to a single GPU with a greater increase at higher orders of accuracy and at lower precision.

GPU acceleration for general conservation equations and its application to several engineering problems

Presented is a general method for conservation equations called SHLL (split HLL) applied using Graphics Processing Unit (GPU) acceleration. The SHLL method is a purely vector-split approximation of the classical HLL method (Harten et al., 1983 [1]) which assumes the presence of local wave propagation in the algebraic derivation of fluxes across cell surfaces. The conventional HLL flux expression terms are interface-split and wave propagation velocities estimated (where required) based on local conditions. Due to the highly local nature of the SHLL fluxes, the scheme is very efficiently applied to GPU computation since the flux, initialisation and update phases of the computation are all vectorized processes. The SHLL scheme is applied to GPU computating using Nvidia’s CUDA package. Numerical schemes are presented for solutions to the general transport (convection–diffusion) equation, Euler Equations and Shallow Water Equations with results presented for several benchmark gas and shallow water flow engineering problems. Computational times are compared between high-end GPU (Nvidia C1060) and CPU (Intel Xeon 3.0 GHz, 32 MB cache) systems with reported speedups of over 67 times when applied to two dimensional simulations with multi-million cell numbers.

Accelerating epistasis analysis in human genetics with consumer graphics hardware

BackgroundHuman geneticists are now capable of measuring more than one million DNA sequence variations from across the human genome. The new challenge is to develop computationally feasible methods capable of analyzing these data for associations with common human disease, particularly in the context of epistasis. Epistasis describes the situation where multiple genes interact in a complex non-linear manner to determine an individual's disease risk and is thought to be ubiquitous for common diseases. Multifactor Dimensionality Reduction (MDR) is an algorithm capable of detecting epistasis. An exhaustive analysis with MDR is often computationally expensive, particularly for high order interactions. This challenge has previously been met with parallel computation and expensive hardware. The option we examine here exploits commodity hardware designed for computer graphics. In modern computers Graphics Processing Units (GPUs) have more memory bandwidth and computational capability than Central Processing Units (CPUs) and are well suited to this problem. Advances in the video game industry have led to an economy of scale creating a situation where these powerful components are readily available at very low cost. Here we implement and evaluate the performance of the MDR algorithm on GPUs. Of primary interest are the time required for an epistasis analysis and the price to performance ratio of available solutions.FindingsWe found that using MDR on GPUs consistently increased performance per machine over both a feature rich Java software package and a C++ cluster implementation. The performance of a GPU workstation running a GPU implementation reduces computation time by a factor of 160 compared to an 8-core workstation running the Java implementation on CPUs. This GPU workstation performs similarly to 150 cores running an optimized C++ implementation on a Beowulf cluster. Furthermore this GPU system provides extremely cost effective performance while leaving the CPU available for other tasks. The GPU workstation containing three GPUs costs $2000 while obtaining similar performance on a Beowulf cluster requires 150 CPU cores which, including the added infrastructure and support cost of the cluster system, cost approximately $82,500.ConclusionGraphics hardware based computing provides a cost effective means to perform genetic analysis of epistasis using MDR on large datasets without the infrastructure of a computing cluster.

Teraflop per second gravitational lensing ray-shooting using graphics processing units

Gravitational lensing calculation using a direct inverse ray-shooting approach is a computationally expensive way to determine magnification maps, caustic patterns, and light-curves (e.g. as a function of source profile and size). However, as an easily parallelisable calculation, gravitational ray-shooting can be accelerated using programmable graphics processing units (GPUs). We present our implementation of inverse ray-shooting for the NVIDIA G80 generation of graphics processors using the NVIDIA Compute Unified Device Architecture (CUDA) software development kit. We also extend our code to multiple GPU systems, including a 4-GPU NVIDIA S1070 Tesla unit. We achieve sustained processing performance of 182 Gflop/s on a single GPU, and 1.28 Tflop/s using the Tesla unit. We demonstrate that billion-lens microlensing simulations can be run on a single computer with a Tesla unit in timescales of order a day without the use of a hierarchical tree-code.