Consumer Graphics Processing Units Research Articles

Enfoque de alta granularidad para el procesamiento de paquetes en red para aplicaciones tolerantes a la latencia con CUDA (Corvyd)

En este artículo, estamos interesados en investigar el procesamiento por lotes basado en capas para la inspección de paquetes de red en paralelo. Estudios anteriores de inspección de paquetes basada en GPU [1], [2] y [3] lograron ganancias de rendimiento a través de algunas innovaciones, los más importantes se basaron en la latencia de acceso a la memoria ocultando [4] que no es monopolizado por los sistemas GPU. Los sistemas de procesamiento de paquetes en cualquier hardware también deben usar alguna forma de bypass del kernel para evitar la sobrecarga asociada con las pilas de red de propósito general [4] y [6]. Estos estudios producen un rendimiento dramáticamente mejor que sus predecesores. Por el contrario, el enfoque de este proyecto se limita estrictamente a un conjunto de mejoras algorítmicas. El procesamiento por lotes es común a cualquier implementación paralela de alto rendimiento. Este proyecto propone explorar un procesamiento más granular basado en capas que divide los paquetes a través de múltiples rondas de programación para maximizar la homogeneidad del lote y minimizar la divergencia de la GPU. Esto aumentará significativamente la sobrecarga para procesar un solo paquete, además de aumentar la latencia a medida que se requerirá más almacenamiento en búfer, pero este enfoque tiene el potencial de mejorar el rendimiento en cargas de trabajo de IP altamente erráticas, donde el trabajo previo ha favorecido fuertemente las cargas de trabajo uniformes como ejemplos de PoC minimizados para representar sistemas que probablemente tendrían un bajo rendimiento en el campo [4],[5], y [6].

Heterogeneous computing for epidemiological model fitting and simulation

BackgroundOver the last years, substantial effort has been put into enhancing our arsenal in fighting epidemics from both technological and theoretical perspectives with scientists from different fields teaming up for rapid assessment of potentially urgent situations. This paper focusses on the computational aspects of infectious disease models and applies commonly available graphics processing units (GPUs) for the simulation of these models. However, fully utilizing the resources of both CPUs and GPUs requires a carefully balanced heterogeneous approach.ResultsThe contribution of this paper is twofold. First, an efficient GPU implementation for evaluating a small-scale ODE model; here, the basic S(usceptible)-I(nfected)-R(ecovered) model, is discussed. Second, an asynchronous particle swarm optimization (PSO) implementation is proposed where batches of particles are sent asynchronously from the host (CPU) to the GPU for evaluation. The ultimate goal is to infer model parameters that enable the model to correctly describe observed data. The particles of the PSO algorithm are candidate parameters of the model; finding the right one is a matter of optimizing the likelihood function which quantifies how well the model describes the observed data. By employing a heterogeneous approach, in which both CPU and GPU are kept busy with useful work, speedups of 10 to 12 times can be achieved on a moderate machine with a high-end consumer GPU as compared to a high-end system with 32 CPU cores.ConclusionsUtilizing GPUs for parameter inference can bring considerable increases in performance using average host systems with high-end consumer GPUs. Future studies should evaluate the benefit of using newer CPU and GPU architectures as well as applying this method to more complex epidemiological scenarios.

Rapid hologram generation utilizing layer-based approach and graphic rendering for realistic three-dimensional image reconstruction by angular tiling

An approach of rapid hologram generation for the realistic three-dimensional (3-D) image reconstruction based on the angular tiling concept is proposed, using a new graphic rendering approach integrated with a previously developed layer-based method for hologram calculation. A 3-D object is simplified as layered cross-sectional images perpendicular to a chosen viewing direction, and our graphics rendering approach allows the incorporation of clear depth cues, occlusion, and shading in the generated holograms for angular tiling. The combination of these techniques together with parallel computing reduces the computation time of a single-view hologram for a 3-D image of extended graphics array resolution to 176 ms using a single consumer graphics processing unit card.

“Red Tweezers”: Fast, customisable hologram generation for optical tweezers

Holographic Optical Tweezers (HOT) are a versatile way of manipulating microscopic particles in 3D. However, their ease of use has been hampered by the computational load of calculating the holograms, resulting in an unresponsive system. We present a program for generating these holograms on a consumer Graphics Processing Unit (GPU), coupled to an easy-to-use interface in LabVIEW (National Instruments). This enables a HOT system to be set up without writing any additional code, as well as providing a platform enabling the fast generation of other holograms. The GPU engine calculates holograms over 300 times faster than the same algorithm running on a quad core CPU. The hologram algorithm can be altered on-the-fly without recompiling the program, allowing it to be used to control Spatial Light Modulators in any situation where the hologram can be calculated in a single pass. The interface has also been rewritten to take advantage of new features in LabVIEW 2010. It is designed to be easily modified and extended to integrate with hardware other than our own. Program summaryProgram title: Red Tweezers.Catalogue identifier: AEQH_v1_0.Program summary URL:http://cpc.cs.qub.ac.uk/summaries/AEQH_v1_0.html.Program obtainable from: CPC Program Library, Queen’s University, Belfast, N. Ireland.Licensing provisions: GNU General Public License.No. of lines in distributed program, including test data, etc.: 79147.No. of bytes in distributed program, including test data, etc.: 11130332.Distribution format: tar.gz.Programming language: LabVIEW 2010, C++, OpenGL Shader Language.Computer: Intel-based personal computer, nVidia or AMD graphics card supporting OpenGL 2.4.Operating system: Microsoft Windows XP or later.Has the code been vectorised or parallelised?: Designed for GPUs.RAM: 2 Gb (highly dependent on video camera).Classification: 18.External routines: OpenGL, National Instruments Vision Development Module.Nature of problem:This program controls a holographic optical tweezers instrument, including GPU-accelerated rendering of holograms, monitoring the video feed and presenting a user-friendly interface to manipulate particles.Solution method:An extendable LabVIEW user interface, including a plugin architecture, is implemented to provide user-friendly control. The program also contains a component that accelerates the necessary digital holography step by rendering patterns using OpenGL shaders.Restrictions:The rendering engine is single-pass, i.e. iterative Fourier transform algorithms are not accelerated on the graphics card.Additional comments:The rendering engine is a separately-compiled executable controlled via UDP and can be used for other purposes. It allows simple OpenGL shaders to be used to render functions, without writing boilerplate code.Running time:This instrument control program is intended to run for as long as the experiment requires, over days if necessary. It can be re-started without losing most of its state information.

Rapid computation of sodium bioscales using gpu‐accelerated image reconstruction

AbstractQuantitative sodium magnetic resonance imaging permits noninvasive measurement of the tissue sodium concentration (TSC) bioscale in the brain. Computing the TSC bioscale requires reconstructing and combining multiple datasets acquired with a non‐Cartesian acquisition that highly oversamples the center of k‐space. Even with an optimized implementation of the algorithm to compute TSC, the overall processing time exceeds the time required to collect data from the human subject. Such a mismatch presents a challenge for sustained sodium imaging to avoid a growing data backlog and provide timely results. The most computationally intensive portions of the TSC calculation have been identified and accelerated using a consumer graphics processing unit (GPU) in addition to a conventional central processing unit (CPU). A recently developed data organization technique called Compact Binning was used along with several existing algorithmic techniques to maximize the scalability and performance of these computationally intensive operations. The resulting GPU+CPU TSC bioscale calculation is more than 15 times faster than a CPU‐only implementation when processing 256 × 256 × 256 data and 2.4 times faster when processing 128 × 128 × 128 data. This eliminates the possibility of a data backlog for quantitative sodium imaging. The accelerated quantification technique is suitable for general three‐dimensional non‐Cartesian acquisitions and may enable more sophisticated imaging techniques that acquire even more data to be used for quantitative sodium imaging. © 2013 Wiley Periodicals, Inc. Int J Imaging Syst Technol, 23, 29–35, 2013.

Pattern-recognition system, designed on GPU, for discriminating between injured normal and pathological knee cartilage

The aim was to design a pattern-recognition (PR) system for discriminating between normal and pathological knee articular cartilage of the medial femoral (MFC) and tibial condyles (MTC). The data set comprised segmented regions of interest (ROIs) from coronal and sagittal 3-T magnetic resonance images of the MFC and MTC cartilage of young patients, 28 with abnormality-free knee and 16 with pathological findings. The PR system was designed employing the probabilistic neural network classifier, textural features from the segmented ROIs and the leave-one-out evaluation method, while the PR system's precision to “unseen” data was assessed by employing the external cross-validation method. Optimal system design was accomplished on a consumer graphics processing unit (GPU) using Compute Unified Device Architecture parallel programming. PR system design on the GPU required about 3.5min against 15h on a CPU-based system. Highest classification accuracies for the MFC and MTC cartilages were 93.2% and 95.5%, and accuracies to “unseen” data were 89% and 86%, respectively. The proposed PR system is housed in a PC, equipped with a consumer GPU, and it may be easily retrained when new verified data are incorporated in its repository and may be of value as a second-opinion tool in a clinical environment.

Porting marine ecosystem model spin-up using transport matrices to GPUs

Abstract. We have ported an implementation of the spin-up for marine ecosystem models based on transport matrices to graphics processing units (GPUs). The original implementation was designed for distributed-memory architectures and uses the Portable, Extensible Toolkit for Scientific Computation (PETSc) library that is based on the Message Passing Interface (MPI) standard. The spin-up computes a steady seasonal cycle of ecosystem tracers with climatological ocean circulation data as forcing. Since the transport is linear with respect to the tracers, the resulting operator is represented by matrices. Each iteration of the spin-up involves two matrix-vector multiplications and the evaluation of the used biogeochemical model. The original code was written in C and Fortran. On the GPU, we use the Compute Unified Device Architecture (CUDA) standard, a customized version of PETSc and a commercial CUDA Fortran compiler. We describe the extensions to PETSc and the modifications of the original C and Fortran codes that had to be done. Here we make use of freely available libraries for the GPU. We analyze the computational effort of the main parts of the spin-up for two exemplar ecosystem models and compare the overall computational time to those necessary on different CPUs. The results show that a consumer GPU can compete with a significant number of cluster CPUs without further code optimization.

High-performance template tracking

Tracking systems are important in computer vision, with applications in video surveillance, human computer interfaces (HCI), etc. Consumer graphics processing units (GPUs) have experienced an extraordinary evolution in both computing performance and programmability, leading to a greater use of the GPU for non-rendering applications, such as image processing and computer vision tasks. In this work we show an effective particle filtering implementation for real-time template tracking based on the use of a graphics card as a streaming architecture in a translation-rotation-scale model.

Performance models for asynchronous data transfers on consumer Graphics Processing Units

Graphics Processing Units (GPU) have impressively arisen as general-purpose coprocessors in high performance computing applications, since the launch of the Compute Unified Device Architecture (CUDA). However, they present an inherent performance bottleneck in the fact that communication between two separate address spaces (the main memory of the CPU and the memory of the GPU) is unavoidable. The CUDA Application Programming Interface (API) provides asynchronous transfers and streams, which permit a staged execution, as a way to overlap communication and computation. Nevertheless, a precise manner to estimate the possible improvement due to overlapping does not exist, neither a rule to determine the optimal number of stages or streams in which computation should be divided. In this work, we present a methodology that is applied to model the performance of asynchronous data transfers of CUDA streams on different GPU architectures. Thus, we illustrate this methodology by deriving expressions of performance for two different consumer graphic architectures belonging to the more recent generations. These models permit programmers to estimate the optimal number of streams in which the computation on the GPU should be broken up, in order to obtain the highest performance improvements. Finally, we have checked the suitability of our performance models with three applications based on codes from the CUDA Software Development Kit (SDK) with successful results.

ARKCoS: artifact-suppressed accelerated radial kernel convolution on the sphere

We describe a hybrid Fourier/direct space convolution algorithm for compact radial (azimuthally symmetric) kernels on the sphere. For high resolution maps covering a large fraction of the sky, our implementation takes advantage of the inexpensive massive parallelism afforded by consumer graphics processing units (GPUs). Applications involve modeling of instrumental beam shapes in terms of compact kernels, computation of fine-scale wavelet transformations, and optimal filtering for the detection of point sources. Our algorithm works for any pixelization where pixels are grouped into isolatitude rings. Even for kernels that are not bandwidth limited, ringing features are completely absent on an ECP grid. We demonstrate that they can be highly suppressed on the popular HEALPix pixelization, for which we develop a freely available implementation of the algorithm. As an example application, we show that running on a high-end consumer graphics card our method speeds up beam convolution for simulations of a characteristic Planck high frequency instrument channel by two orders of magnitude compared to the commonly used HEALPix implementation on one CPU core while maintaining at typical a fractional RMS accuracy of about 1 part in 10^5.

EpiGPU: exhaustive pairwise epistasis scans parallelized on consumer level graphics cards

Hundreds of genome-wide association studies have been performed over the last decade, but as single nucleotide polymorphism (SNP) chip density has increased so has the computational burden to search for epistasis [for n SNPs the computational time resource is O(n(n-1)/2)]. While the theoretical contribution of epistasis toward phenotypes of medical and economic importance is widely discussed, empirical evidence is conspicuously absent because its analysis is often computationally prohibitive. To facilitate resolution in this field, tools must be made available that can render the search for epistasis universally viable in terms of hardware availability, cost and computational time. By partitioning the 2D search grid across the multicore architecture of a modern consumer graphics processing unit (GPU), we report a 92× increase in the speed of an exhaustive pairwise epistasis scan for a quantitative phenotype, and we expect the speed to increase as graphics cards continue to improve. To achieve a comparable computational improvement without a graphics card would require a large compute-cluster, an option that is often financially non-viable. The implementation presented uses OpenCL--an open-source library designed to run on any commercially available GPU and on any operating system. The software is free, open-source, platformindependent and GPU-vendor independent. It can be downloaded from http://sourceforge.net/projects/epigpu/.

3D nonrigid registration via optimal mass transport on the GPU

In this paper, we present a new computationally efficient numerical scheme for the minimizing flow approach for optimal mass transport (OMT) with applications to non-rigid 3D image registration. The approach utilizes all of the gray-scale data in both images, and the optimal mapping from image A to image B is the inverse of the optimal mapping from B to A. Further, no landmarks need to be specified, and the minimizer of the distance functional involved is unique. Our implementation also employs multigrid, and parallel methodologies on a consumer graphics processing unit (GPU) for fast computation. Although computing the optimal map has been shown to be computationally expensive in the past, we show that our approach is orders of magnitude faster then previous work and is capable of finding transport maps with optimality measures (mean curl) previously unattainable by other works (which directly influences the accuracy of registration). We give results where the algorithm was used to compute non-rigid registrations of 3D synthetic data as well as intra-patient pre-operative and post-operative 3D brain MRI datasets.

SU‐GG‐T‐511: Towards Real‐Time Radiation Therapy: Superposition/Convolution at 4fps

Purpose: To enable novel radiation therapy workflows by providing superposition/convolution dose calculation at interactive rates. Method and Materials: The Superposition/Convolution (S/C) dose computation algorithm utilizes ray‐tracing techniques extensively and is well suited to acceleration using consumer graphics processing units (GPU). The standard S/C algorithm was parallelized and implemented using NVIDIA's CUDA GPU programming environment. The basic S/C method generates incident fluence, using a source model, and transports it from the radiation source, using ray‐tracing, to deposit a TERMA grid. Then superposition spreads the dose deposition kernel out by ray‐tracing from each point in the TERMA grid. There are several issues with parallelizing this algorithm. Calculation of the TERMA grid suffers from read‐write conflicts. We solve this by running groups of spatially separated divergent rays. The forward S/C kernel also suffers from read‐write conflicts, but the inverse kernel does not. We reduced memory bandwidth by using a variable step ray‐tracer, independently attenuating each spectral energy bin and caching lookup‐tables. We explored the concept of using volumetric mip‐maps to approximate the ray as a true solid angle. Results: We compared our implementation to Pinnacle3 for low (64×64×64 cube with 4 mm voxels) and high resolution (128×128×128 cube with 2mm voxels) water phantoms using as similar settings as possible. Pinnacle3 (Philips ‐ Madison, WI) was run on a SunFire v250 machine with times of 30.961s and 365.944s respectively. We performed our calculations on a single NVIDIA GeForce 8800 GTX with times of 0.217s and 2.197s respectively. Preliminary experiments using volumetric mip‐maps showed additional performance improvements with minor accuracy loss. Conclusion: We have completed a GPU accelerated superposition/convolution dose engine, providing a substantial performance gain over CPU based implementations — indicating that real time dose computation is feasible with the accuracy levels of the S/C algorithm.

Multiscale and local search methods for real time region tracking with particle filters: local search driven by adaptive scale estimation on GPUs

Tracking systems are important in computervision, with applications in surveillance, human computer interaction, etc. Consumer graphics processing units (GPUs) have experienced an extraordinary evolution in both computing performance and programmability, leading to greater use of the GPU for non-rendering applications. In this work we propose a real-time object tracking algorithm, based on the hybridization of particle filtering (PF) and a multi-scale local search (MSLS) algorithm, presented for both CPU and GPU architectures. The developed system provides successful results in precise tracking of single and multiple targets in monocular video, operating in real-time at 70 frames per second for 640 × 480 video resolutions on the GPU, up to 1,100% faster than the CPU version of the algorithm.