Acceleration-as-a-Service: Exploiting Virtualised GPUs for a Financial Application

How can GPU acceleration be obtained as a service in a cluster? This question has become increasingly significant due to the inefficiency of installing GPUs on all nodes of a cluster. The research reported in this paper is motivated to address the above question by employing rCUDA (remote CUDA), a framework that facilitates Acceleration-as-a-Service (AaaS), such that the nodes of a cluster can request the acceleration of a set of remote GPUs on demand. The rCUDA framework exploits virtualisation and ensures that multiple nodes can share the same GPU. In this paper we test the feasibility of the rCUDA framework on a real-world application employed in the financial risk industry that can benefit from AaaS in the production setting. The results confirm the feasibility of rCUDA and highlight that rCUDA achieves similar performance compared to CUDA, provides consistent results, and more importantly, allows for a single application to benefit from all the GPUs available in the cluster without loosing efficiency.

GPU accelerated MCMC for modelling terrorist activity

The use of graphical processing unit (GPU) parallel processing is becoming a part of mainstream statistical practice. The reliance of Bayesian statistics on Markov Chain Monte Carlo (MCMC) methods makes the applicability of parallel processing not immediately obvious. It is illustrated that there are substantial gains in improved computational time for MCMC and other methods of evaluation by computing the likelihood using GPU parallel processing. Examples use data from the Global Terrorism Database to model terrorist activity in Colombia from 2000 through 2010 and a likelihood based on the explicit convolution of two negative-binomial processes. Results show decreases in computational time by a factor of over 200. Factors influencing these improvements and guidelines for programming parallel implementations of the likelihood are discussed.

Jacobian-free Newton-Krylov methods with GPU acceleration for computing nonlinear ship wave patterns

The nonlinear problem of steady free-surface flow past a submerged source is considered as a case study for three-dimensional ship wave problems. Of particular interest is the distinctive wedge-shaped wave pattern that forms on the surface of the fluid. By reformulating the governing equations with a standard boundary-integral method, we derive a system of nonlinear algebraic equations that enforce a singular integro-differential equation at each midpoint on a two-dimensional mesh. Our contribution is to solve the system of equations with a Jacobian-free Newton-Krylov method together with a banded preconditioner that is carefully constructed with entries taken from the Jacobian of the linearised problem. Further, we are able to utilise graphics processing unit acceleration to significantly increase the grid refinement and decrease the run-time of our solutions in comparison to schemes that are presently employed in the literature. Our approach provides opportunities to explore the nonlinear features of three-dimensional ship wave patterns, such as the shape of steep waves close to their limiting configuration, in a manner that has been possible in the two-dimensional analogue for some time.

GPU accelerated algorithms for computing matrix function vector products with applications to exponential integrators and fractional diffusion

The efficient computation of matrix function vector products has become an important area of research in recent times, driven in particular by two important applications: the numerical solution of fractional partial differential equations and the integration of large systems of ordinary differential equations. In this work we consider a problem that combines these two applications, in the form of a numerical solution algorithm for fractional reaction diffusion equations that after spatial discretisation, is advanced in time using the exponential Euler method. We focus on the efficient implementation of the algorithm on Graphics Processing Units (GPU), as we wish to make use of the increased computational power available with this hardware. We compute the matrix function vector products using the contour integration method in [N. Hale, N. Higham, and L. Trefethen. Computing Aα, log(A), and related matrix functions by contour integrals. SIAM J. Numer. Anal., 46(5):2505–2523, 2008]. Multiple levels of preconditioning are applied to reduce the GPU memory footprint and to further accelerate convergence. We also derive an error bound for the convergence of the contour integral method that allows us to pre-determine the appropriate number of quadrature points. Results are presented that demonstrate the effectiveness of the method for large two-dimensional problems, showing a speedup of more than an order of magnitude compared to a CPU-only implementation.

Efficient asynchronous executions of AMR computations and visualization on a GPU system

Adaptive Mesh Refinement is a method which dynamically varies the spatio-temporal resolution of localized mesh regions in numerical simulations, based on the strength of the solution features. In-situ visualization plays an important role for analyzing the time evolving characteristics of the domain structures. Continuous visualization of the output data for various timesteps results in a better study of the underlying domain and the model used for simulating the domain. In this paper, we develop strategies for continuous online visualization of time evolving data for AMR applications executed on GPUs. We reorder the meshes for computations on the GPU based on the users input related to the subdomain that he wants to visualize. This makes the data available for visualization at a faster rate. We then perform asynchronous executions of the visualization steps and fix-up operations on the CPUs while the GPU advances the solution. By performing experiments on Tesla S1070 and Fermi C2070 clusters, we found that our strategies result in 60% improvement in response time and 16% improvement in the rate of visualization of frames over the existing strategy of performing fix-ups and visualization at the end of the timesteps.

Automatic Data Allocation and Buffer Management for Multi-GPU Machines

Multi-GPU machines are being increasingly used in high-performance computing. Each GPU in such a machine has its own memory and does not share the address space either with the host CPU or other GPUs. Hence, applications utilizing multiple GPUs have to manually allocate and manage data on each GPU. Existing works that propose to automate data allocations for GPUs have limitations and inefficiencies in terms of allocation sizes, exploiting reuse, transfer costs, and scalability. We propose a scalable and fully automatic data allocation and buffer management scheme for affine loop nests on multi-GPU machines. We call it the Bounding-Box-based Memory Manager (BBMM). BBMM can perform at runtime, during standard set operations like union, intersection, and difference, finding subset and superset relations on hyperrectangular regions of array data (bounding boxes). It uses these operations along with some compiler assistance to identify, allocate, and manage data required by applications in terms of disjoint bounding boxes. This allows it to (1) allocate exactly or nearly as much data as is required by computations running on each GPU, (2) efficiently track buffer allocations and hence maximize data reuse across tiles and minimize data transfer overhead, and (3) and as a result, maximize utilization of the combined memory on multi-GPU machines. BBMM can work with any choice of parallelizing transformations, computation placement, and scheduling schemes, whether static or dynamic. Experiments run on a four-GPU machine with various scientific programs showed that BBMM reduces data allocations on each GPU by up to 75% compared to current allocation schemes, yields performance of at least 88% of manually written code, and allows excellent weak scaling.

GPU-based normalized cuts for road extraction using satellite imagery

This paper presents a GPU implementation of normalized cuts for road extraction problem using panchromatic satellite imagery. The roads have been extracted in three stages namely pre-processing, image segmentation and post-processing. Initially, the image is pre-processed to improve the tolerance by reducing the clutter (that mostly represents the buildings, vegetation,. and fallow regions). The road regions are then extracted using the normalized cuts algorithm. Normalized cuts algorithm is a graph-based partitioning `approach whose focus lies in extracting the global impression (perceptual grouping) of an image rather than local features. For the segmented image, post-processing is carried out using morphological operations - erosion and dilation. Finally, the road extracted image is overlaid on the original image. Here, a GPGPU (General Purpose Graphical Processing Unit) approach has been adopted to implement the same algorithm on the GPU for fast processing. A performance comparison of this proposed GPU implementation of normalized cuts algorithm with the earlier algorithm (CPU implementation) is presented. From the results, we conclude that the computational improvement in terms of time as the size of image increases for the proposed GPU implementation of normalized cuts. Also, a qualitative and quantitative assessment of the segmentation results has been projected.

3-D GPU Based Real Time Diffuse Optical Tomographic System

3-Dimensional Diffuse Optical Tomographic (3-D DOT) image reconstruction algorithm is computationally complex and requires excessive matrix computations and thus hampers reconstruction in real time. In this paper, we present near real time 3D DOT image reconstruction that is based on Broyden approach for updating Jacobian matrix. The Broyden method simplifies the algorithm by avoiding re-computation of the Jacobian matrix in each iteration. We have developed CPU and heterogeneous CPU/GPU code for 3D DOT image reconstruction in C and MatLab programming platform. We have used Compute Unified Device Architecture (CUDA) programming framework and CUDA linear algebra library (CULA) to utilize the massively parallel computational power of GPUs (NVIDIA Tesla K20c). The computation time achieved for C program based implementation for a CPU/GPU system for 3 planes measurement and FEM mesh size of 19172 tetrahedral elements is 806 milliseconds for an iteration.

Adaptive executions of hyperbolic block-structured AMR applications on GPU systems

A block-structured adaptive mesh refinement (AMR) technique has been used to obtain numerical solutions for many scientific applications. Some block-structured AMR approaches have focused on forming patches of non-uniform sizes where the size of a patch can be tuned to the geometry of a region of interest. In this paper, we develop strategies for adaptive execution of block-structured AMR applications on GPUs, for hyperbolic directionally split solvers. While effective hybrid execution strategies exist for applications with uniform patches, our work considers efficient execution of non-uniform patches with different workloads. Our techniques include bin-packing work units to load balance GPU computations, adaptive asynchronism between CPU and GPU executions using a knapsack formulation, and scheduling communications for multi-GPU executions. Our experiments with synthetic and real data, for single-GPU and multi-GPU executions, on Tesla S1070 and Fermi C2070 clusters, show that our strategies result in up to a 3.23 speedup in performance over existing strategies.

Computer generated hologram with geometric occlusion using GPU-accelerated depth buffer rasterization for three-dimensional display

We present a method of rapidly producing computer-generated holograms that exhibit geometric occlusion in the reconstructed image. Conceptually, a bundle of rays is shot from every hologram sample into the object volume.We use z buffering to find the nearest intersecting object point for every ray and add its complex field contribution to the corresponding hologram sample. Each hologram sample belongs to an independent operation, allowing us to exploit the parallel computing capability of modern programmable graphics processing units (GPUs). Unlike algorithms that use points or planar segments as the basis for constructing the hologram, our algorithm's complexity is dependent on fixed system parameters, such as the number of ray-casting operations, and can therefore handle complicated models more efficiently. The finite number of hologram pixels is, in effect, a windowing function, and from analyzing the Wigner distribution function of windowed free-space transfer function we find an upper limit on the cone angle of the ray bundle. Experimentally, we found that an angular sampling distance of 0:01' for a 2:66' cone angle produces acceptable reconstruction quality. © 2009 Optical Society of America.

Restauração de imagens de microscopia de força atômica com uso da regularização de Tikhonov via processamento em GPU

A Restauração de Imagens é uma técnica que possui aplicações em várias áreas, por exemplo, medicina, biologia, eletrônica, e outras, onde um dos objetivos da restauração de imagens é melhorar o aspecto final de imagens de amostras que por algum motivo apresentam imperfeições ou borramentos. As imagens obtidas pelo Microscópio de Força Atômica apresentam borramentos causados pela interação de forças entre a ponteira do microscópio e a amostra em estudo. Além disso apresentam ruídos aditivos causados pelo ambiente. Neste trabalho é proposta uma forma de paralelização em GPU de um algoritmo de natureza serial que tem por fim a Restauração de Imagens de Microscopia de Força Atômica baseado na Regularização de Tikhonov.

DistributedCL: middleware de processamento distribuído em GPU com interface da API OpenCL.

Este trabalho apresenta a proposta de um middleware, chamado DistributedCL, que torna transparente o processamento paralelo em GPUs distribuídas. Com o suporte do middleware DistributedCL uma aplicação, preparada para utilizar a API OpenCL, pode executar de forma distribuída, utilizando GPUs remotas, de forma transparente e sem necessidade de alteração ou nova compilação do seu código. A arquitetura proposta para o middleware DistributedCL é modular, com camadas bem definidas e um protótipo foi construído de acordo com a arquitetura, onde foram empregados vários pontos de otimização, incluindo o envio de dados em lotes, comunicação assíncrona via rede e chamada assíncrona da API OpenCL. O protótipo do middleware DistributedCL foi avaliado com o uso de benchmarks disponíveis e também foi desenvolvido o benchmark CLBench, para avaliação de acordo com a quantidade dos dados. O desempenho do protótipo se mostrou bom, superior às propostas semelhantes, tendo alguns resultados próximos do ideal, sendo o tamanho dos dados para transmissão através da rede o maior fator limitante.

Acceleration of an unstructured hybrid mesh RANS solver by porting to GPU architectures

The modern CFD process consists of mesh generation, flow solving and post-processing integrated into an automated workflow. During the last several years we have developed and published research aimed at producing a meshing and geometry editing system, implemented in an end-to-end parallel, scalable manner and capable of automatic handling of large scale, real world applications. The particular focus of this paper is the associated unstructured mesh RANS flow solver and the porting of it to GPU architectures. After briefly describing the solver itself, the special issues associated with porting codes using unstructured data structures are discussed - followed by some application examples. Copyright © 2011 by W.N. Dawes.