Commodity Graphics Processing Units Research Articles

WFA-GPU: gap-affine pairwise read-alignment using GPUs.

Advances in genomics and sequencing technologies demand faster and more scalable analysis methods that can process longer sequences with higher accuracy. However, classical pairwise alignment methods, based on dynamic programming (DP), impose impractical computational requirements to align long and noisy sequences like those produced by PacBio and Nanopore technologies. The recently proposed wavefront alignment (WFA) algorithm paves the way for more efficient alignment tools, improving time and memory complexity over previous methods. However, high-performance computing (HPC) platforms require efficient parallel algorithms and tools to exploit the computing resources available on modern accelerator-based architectures. This paper presents WFA-GPU, a GPU (graphics processing unit)-accelerated tool to compute exact gap-affine alignments based on the WFA algorithm. We present the algorithmic adaptations and performance optimizations that allow exploiting the massively parallel capabilities of modern GPU devices to accelerate the alignment computations. In particular, we propose a CPU-GPU co-design capable of performing inter-sequence and intra-sequence parallel sequence alignment, combining a succinct WFA-data representation with an efficient GPU implementation. As a result, we demonstrate that our implementation outperforms the original multi-threaded WFA implementation by up to 4.3× and up to 18.2× when using heuristic methods on long and noisy sequences. Compared to other state-of-the-art tools and libraries, the WFA-GPU is up to 29× faster than other GPU implementations and up to four orders of magnitude faster than other CPU implementations. Furthermore, WFA-GPU is the only GPU solution capable of correctly aligning long reads using a commodity GPU. WFA-GPU code and documentation are publicly available at https://github.com/quim0/WFA-GPU.

Validation of organ dose calculations with PyMCGPU-IR in realistic interventional set-ups.

Interventional radiology procedures are associated with high skin dose exposure. The 2013/59/EURATOM Directive establishes that the equipment used for interventional radiology must have a device or a feature informing the practitioner of relevant parameters for assessing patient dose at the end of the procedure. This work presents and validates PyMCGPU-IR, a patient dose monitoring tool for interventional cardiology and radiology procedures based on MC-GPU. MC-GPU is a freely available Monte Carlo (MC) code of photon transport in a voxelized geometry which uses the computational power of commodity Graphics Processing Unit cards (GPU) to accelerate calculations. PyMCGPU-IR was validated against two different experimental set-ups. The first one consisted of skin dose measurements for different beam angulations on an adult Rando Alderson anthropomorphic phantom. The second consisted of organ dose measurements in three clinical procedures using the Rando Alderson phantom. The results obtained for the skin dose measurements show differences below 6%. For the clinical procedures the differences are within 20% for most cases. PyMCGPU-IR offers both, high performance and accuracy for dose assessment when compared with skin and organ dose measurements. It also allows the calculation of dose values at specific positions and organs, the dose distribution and the location of the maximum doses per organ. In addition, PyMCGPU-IR overcomes the time limitations of CPU-based MC codes.

A Range and Performance Optimized Version of the Computer-Aided Speckle Interferometry Algorithm for Real-Time Displacement-Strain Field Monitoring

This work presents an optimized implementation of the Computer-Aided Speckle Interferometry algorithm which enables full-field determination of displacements and strains on commodity Graphics Processing Units at high resolution and frame rates. By combining careful control of the average speckle size in a laser speckle pattern with a simple sampling rate conversion scheme, a compact representation of the optical speckle is achieved. This allows for optimal use of Graphics Processing Unit architecture with robust range extension. The optimal mapping of the Computer-Aided Speckle Interferometry algorithm to Graphics Processing Unit architecture is shown in detail, and a straightforward method for disambiguating large displacements is illustrated. Lastly, this paper demonstrates a two-step subimage-tapering modification to the original algorithm that enables robust range enhancement while maintaining resolution. Results from numerical simulations on synthetic speckle patterns are shown, and runtime performance metrics are provided, with performance ranging up to 60 frames per second in some cases. The method is suitable for interactive experimental mechanics research, process and testing or any application where real-time high-resolution displacement-strain monitoring is needed. A .NET Framework class library enabling the incorporation of the algorithm into 3rd -party applications is available for download.

Parallel and Distributed Computing for Anomaly Detection From Hyperspectral Remote Sensing Imagery

Anomaly detection from remote sensing images is to detect pixels whose spectral signatures are different from their background. Anomalies are often man-made targets. With such target signatures being unknown, anomaly detection has many important applications, such as water quality monitoring, crop stress surveying, and law enforcement-related uses, where prior information of targets is often unavailable. The key to success is accurate background modeling. Anomaly detection from remote sensing images is challenging because spatial coverage is very large and the background is highly heterogeneous. For pixel-based anomaly detection, computing cost in background modeling and a spatial-convolution-type detection process is very expensive. Thus, parallel and distributed computing is critical in reducing execution time, which can fit the need for real-time or near real-time detection from airborne and spaceborne platforms in support of immediate decision-making. This article reviews the recent advances in anomaly detection from hyperspectral remote sensing images and their implementation using parallel and distributed systems. The classical methods, i.e., the Reed-Xiaoli (RX) algorithm and its variants, including its real-time processing version, are illustrated in commodity graphic processing units (GPUs), cloud, and field-programmable gate array (FPGA) implementations. Practical issues and future development trends are also discussed.

Open GL–Open CL Solar Radiation Pressure Modeling with Time-Varying Spacecraft Geometries

A method for the fast computation of spacecraft force and torque due to solar radiation pressure (SRP) is presented. A faceted model is employed that tracks which elements of a time-varying geometry are exposed to the sunlight, but sunlight reflections are not modeled. The method uses the highly parallel execution capabilities of commodity graphics processing unit (GPU) and the Open Graphics Library (OpenGL) and Open Compute Language (OpenCL) to render a spacecraft mesh on the GPU. A custom-developed OpenGL render pipeline computes the per-model facet SRP forces and torques that are summed on the GPU before the resultant spacecraft force and torque vectors are copied back to the CPU bound process. The process is validated on spherical and cubic test shapes. The evaluation accommodates spacecraft self-shadowing and is capable of accounting for arbitrary spacecraft articulation. Material properties are encoded with the model to provide realistic specular, diffuse, and absorption surface light interactions. Numerical simulations illustrate the impact of geometric fidelity and articulated surfaces. The faceted OpenGL-OpenCL method is up to an order of magnitude faster on integrated GPU hardware than high-end graphics card as the process is not demanding on the GPU and benefits from the fast memory transfer of on-chip processors.

Intelligent Classification of Large-Scale Remotely Sensed Hyperspectral Images using Multi-GPU Computing

Image classification is one of the most popular tasks used to analyze remote sensing signatures (RSS) of a geographical region from remotely sensed hyperspectral images. However, the high dimensionality of such hyperspectral images raises a series of new challenges. In this paper, a new approach for real-time intelligent classification of large-scale hyperspectral imagery which aggregates Fuzzy logic and the fused weighted order statistic (WOS) with the minimum distance to mean (MDM) techniques using commodity graphics processing units (GPUs) is addressed. Within this context, intelligent image processing methods are algorithmically adapted via parallel computing techniques and efficiently implemented in two NVIDIA Tesla C2075 GPUs. Experimental results demonstrate how such unification reduces drastically the computational load of the real-world hyperspectral classification tasks resulting in efficient numerical algorithms suitable for real-time multi-GPU-adapted implementation.

High-Performance Homomorphic Matrix Completion on Multiple GPUs

In various applications such as trajectory tracking in mobile social networks and online recommendation systems, the massive raw data are often incomplete due to various unpredictable or unavoidable reasons. Matrix completion algorithms are effective for reconstructing two-dimensional data, but sending raw data containing personal, sensitive information to cloud computing nodes for matrix completion may lead to privacy exposure issue. The homomorphic matrix completion is a promising approach to perform matrix completion while preserving privacy. However, CPU-based homomorphic matrix completion has low performance, making it impractical to process multiple or large-scale data completion tasks in real-time. In this paper, we propose a high-performance homomorphic matrix completion scheme by exploiting commodity GPUs (Graphics Processing Units) that are widely available in HPC servers and cloud computing nodes. First, we design and implement a baseline GPU-based homomorphic matrix completion, and propose techniques to optimize memory accesses, GPU utilization, and communications. Second, we propose a shard mode for large-scale matrices exceeding GPU memory capacity. Third, we propose a multi-GPU mode to fully utilize multiple GPUs in computing nodes. Experiment results show that the proposed scheme is both fast and accurate. On matrices of varying sizes, the proposed scheme running on a single Tesla V100 GPU achieves up to 116.23× speedups over the CPU MATLAB implementation running on dual Xeon CPUs. The multi-GPU mode achieves up to 1.84× speedups on two GPUs versus on a single GPU. For large-scale matrices, the shard mode achieves up to 174.92× speedups on a single GPU over the CPU MATLAB implementation on two CPUs, and further achieves up to 1.35× speedups when running on two GPUs using the multi-GPU mode.

ZeroKernel: Secure Context-isolated Execution on Commodity GPUs

In the last decade, the dedicated graphics processing unit (GPU) has emerged as an architecture for high-performance computing workloads. Recently, researchers have also focused on the isolation property of a dedicated GPU and suggested GPU-based secure computing environments with several promising applications. However, despite the security analysis conducted by the prior studies, it has been unclear whether a dedicated GPU can be leveraged as a secure processor in the presence of a kernel-privileged attacker. In this paper, we first demonstrate the security of dedicated GPUs through comprehensive studies on context information for GPU execution. The paper shows that a kernel-privileged attacker can manipulate the GPU contexts to redirect memory accesses or execute arbitrary GPU codes on the running GPU kernel. Based on the security analysis, this paper proposes a new on-chip execution model for the dedicated GPU and a novel defense mechanism supporting the security of the on-chip execution. With comprehensive evaluation, the paper assures that the proposed solutions effectively isolate sensitive data in on-chip storages and defend against known attack vectors from a privileged attacker, supporting that the commodity GPUs can be leveraged as a secure processor.

GPU Parallel Implementation of Spatially Adaptive Hyperspectral Image Classification

Image classification is a very important tool for remotely sensed hyperspectral image processing. Techniques able to exploit the rich spectral information contained in the data, as well as its spatial-contextual information, have shown success in recent years. Due to the high dimensionality of hyperspectral data, spectral-spatial classification techniques are quite demanding from a computational viewpoint. In this paper, we present a computationally efficient parallel implementation for a spectral-spatial classification method based on spatially adaptive Markov random fields (MRFs). The method learns the spectral information from a sparse multinomial logistic regression classifier, and the spatial information is characterized by modeling the potential function associated with a weighted MRF as a spatially adaptive vector total variation function. The parallel implementation has been carried out using commodity graphics processing units (GPUs) and the NVIDIA's Compute Unified Device Architecture. It optimizes the work allocation and input/output transfers between the central processing unit and the GPU, taking full advantages of the computational power of GPUs as well as the high bandwidth and low latency of shared memory. As a result, the algorithm exploits the massively parallel nature of GPUs to achieve significant acceleration factors (higher than 70x) with regards to the serial and multicore versions of the same classifier on an NVIDIA Tesla K20C platform.

Parallel Implementation of Spatial–Spectral Endmember Extraction on Graphic Processing Units

The identification of pure spectral signatures (endmembers) in remotely sensed hyperspectral images has traditionally focused on the spectral information alone. Recently, techniques such as the spatial–spectral endmember extraction (SSEE) have incorporated both the spectral and the spatial information contained in the scene. Since hyperspectral images contain very detailed information in the spatial and spectral domain, the integration of these two sources of information generally comes with a significant increase in computational complexity. In this paper, we develop a new computationally efficient implementation of SSEE using commodity graphics processing units (GPUs). The relevance of GPUs comes from their very low cost, compact size, and the possibility to obtain significant acceleration factors by exploiting properly the GPU hardware architecture. Our experimental results, focused on evaluating the candidate endmembers produced by SSEE and also the computational performance of the GPU implementation, indicated significant acceleration factors that allow exploiting the SSEE method in computationally efficient fashion.

Execution of a parallel edge-based Navier–Stokes solver on commodity graphics processor units

ABSTRACTThe implementation of an edge-based three-dimensional Reynolds Average Navier–Stokes solver for unstructured grids able to run on multiple graphics processing units (GPUs) is presented. Loops over edges, which are the most time-consuming part of the solver, have been written to exploit the massively parallel capabilities of GPUs. Non-blocking communications between parallel processes and between the GPU and the central processor unit (CPU) have been used to enhance code scalability. The code is written using a mixture of C++ and OpenCL, to allow the execution of the source code on GPUs. The Message Passage Interface (MPI) library is used to allow the parallel execution of the solver on multiple GPUs. A comparative study of the solver parallel performance is carried out using a cluster of CPUs and another of GPUs. It is shown that a single GPU is up to 64 times faster than a single CPU core. The parallel scalability of the solver is mainly degraded due to the loss of computing efficiency of the GPU when the size of the case decreases. However, for large enough grid sizes, the scalability is strongly improved. A cluster featuring commodity GPUs and a high bandwidth network is ten times less costly and consumes 33% less energy than a CPU-based cluster with an equivalent computational power.

Enhancement and Edge-Preserving Denoising: An OpenCL-Based Approach for Remote Sensing Imagery

Image enhancement and edge-preserving denoising are relevant steps before classification or other postprocessing techniques for remote sensing images. However, multisensor array systems are able to simultaneously capture several low-resolution images from the same area on different wavelengths, forming a high spatial/spectral resolution image and raising a series of new challenges. In this paper, an open computing language based parallel implementation approach is presented for near real-time enhancement based on Bayesian maximum entropy (BME), as well as an edge-preserving denoising algorithm for remote sensing imagery, which uses the local linear Stein's unbiased risk estimate (LLSURE). BME was selected for its results on synthetic aperture radar image enhancement, whereas LLSURE has shown better noise removal properties than other commonly used methods. Within this context, image processing methods are algorithmically adapted via parallel computing techniques and efficiently implemented using CPUs and commodity graphics processing units (GPUs). Experimental results demonstrate the reduction of computational load of real-world image processing for near real-time GPU adapted implementation.

GPU classification of remote-sensing images using kernel ELM and extended morphological profiles

ABSTRACTNowadays, the use of hyperspectral sensors has been extended to a variety of applications such as the classification of remote-sensing images. Recently, a spectral–spatial classification scheme (ELM-EMP) based on Extreme Learning Machine (ELM) and Extended Morphological Profiles (EMPs) computed using Principal Component Analysis (PCA) and morphological operations has been introduced. In this work, an efficient implementation of this scheme over commodity Graphics Processing Units (GPUs) is shown. Additionally, several techniques and optimizations are introduced to improve the accuracy of the classification. In particular, a scheme using an ELM classifier based on kernels (KELM) and EMP is presented (KELM-EMP). Similar schemes adding a spatial regularization process (KELM-EMP-S and ELM-EMP-S) are also proposed. Moreover, two PCA algorithms have been compared in both accuracy and speed terms. Regarding the GPU projection, different techniques and optimizations have been applied such as the use of optimized Compute Unified Device Architecture (CUDA) libraries or a block-asynchronous execution technique. As a result, the accuracy obtained by the two proposed schemes (ELM-EMP-S and KELM-EMP-S) is better than for the original scheme ELM-EMP and the execution time has been significantly reduced.

GPU Implementation of Spatial–Spectral Preprocessing for Hyperspectral Unmixing

Spectral unmixing pursues the identification of spectrally pure constituents, called endmembers , and their corresponding abundances in each pixel of a hyperspectral image. Most unmixing techniques have focused on the exploitation of spectral information alone. Recently, some techniques have been developed to take advantage of the complementary information provided by the spatial correlation of the pixels in the image. Computational complexity represents a major problem in these spatial–spectral techniques, as hyperspectral images contain very rich information in both the spatial and spectral domains. In this letter, we develop a computationally efficient implementation of a spatial–spectral processing algorithm that has been successfully applied prior to the spectral unmixing of the hyperspectral data. Our implementation has been optimized for the commodity graphics processing units (GPUs) and is evaluated (using both synthetic and real data) using different GPU architectures. Significant speedups can be achieved when processing hyperspectral images of different sizes. This allows for the inclusion of the proposed parallel preprocessing module in a full hyperspectral unmixing chain able to operate in real time.

Efficient implementation of morphological index for building/shadow extraction from remotely sensed images

Morphological building index (MBI) and morphological shadow index (MSI) are recently developed techniques that aim at automatically detect buildings/shadows using high-resolution remotely sensed imagery. The traditional mathematical morphology operations are usually time-consuming as they are based on the consideration of a wide range of image-object properties, such as brightness, contrast, shapes, sizes, and in the application of series of repeated transformations (e.g., classical opening and closing operators). In the case of MBI and MSI, the computational complexity is also increased due to the use of multiscale and multidirectional morphological operators. In this paper, we provide a computationally efficient implementation of MBI and MSI algorithms which is specifically developed for commodity graphic processing units using NVIDIA CUDA. We perform the evaluation of the parallel version of the algorithms using two different NVIDIA architectures and three widely used hyperspectral data sets. Experimental results show that the computational burden introduced when considering multidirectional morphological operators can be almost completely removed by the developed implementations.

GPUvm: GPU Virtualization at the Hypervisor

Graphic processing units (GPUs) provide a massively-parallel computational power and encourage the use of general-purpose computing on GPUs (GPGPU). The distinguished design of <i>discrete GPUs</i> helps them to provide the high throughput, scalability, and energy efficiency needed for GPGPU applications. Despite the previous study on GPU virtualization, the tradeoffs between the virtualization approaches remain unclear, because of a lack of designs for or quantitative evaluations of the hypervisor-level virtualization for discrete GPUs. Shedding light on these tradeoffs and the technical requirements for the hypervisor-level virtualization would facilitate the development of an appropriate GPU virtualization solution. <italic/> <inline-formula><tex-math notation="LaTeX"> $\sf{GPUvm}$</tex-math> </inline-formula> <italic/> , which is an open architecture for hypervisor-level GPU virtualization with a particular emphasis on using the Xen hypervisor, is presented in this paper. <inline-formula><tex-math notation="LaTeX"> $\sf{GPUvm}$</tex-math> </inline-formula> offers three virtualization modes: the full-, naive para-, and high-performance para-virtualization. <inline-formula><tex-math notation="LaTeX">$\sf{GPUvm}$</tex-math></inline-formula> exposes low- and high-level interfaces such as memory-mapped I/O and DRM APIs to the guest virtual machines (VMs). Our experiments using a relevant commodity GPU showed that <inline-formula><tex-math notation="LaTeX">$\sf{GPUvm}$</tex-math></inline-formula> incurs different overheads as the level of the exposed interfaces is changed. The results also showed that a coarse-grained fairness on the GPU among multiple VMs can be achieved using GPU scheduling.

Parallel Hyperspectral Coded Aperture for Compressive Sensing on GPUs

The application of compressive sensing (CS) to hyperspectral images is an active area of research over the past few years, both in terms of the hardware and the signal processing algorithms. However, CS algorithms can be computationally very expensive due to the extremely large volumes of data collected by imaging spectrometers, a fact that compromises their use in applications under real-time constraints. This paper proposes four efficient implementations of hyperspectral coded aperture (HYCA) for CS, two of them termed P-HYCA and P-HYCA-FAST and two additional implementations for its constrained version (CHYCA), termed P-CHYCA and P-CHYCA-FAST on commodity graphics processing units (GPUs). HYCA algorithm exploits the high correlation existing among the spectral bands of the hyperspectral data sets and the generally low number of endmembers needed to explain the data, which largely reduces the number of measurements necessary to correctly reconstruct the original data. The proposed P-HYCA and P-CHYCA implementations have been developed using the compute unified device architecture (CUDA) and the cuFFT library. Moreover, this library has been replaced by a fast iterative method in the P-HYCA-FAST and P-CHYCA-FAST implementations that leads to very significant speedup factors in order to achieve real-time requirements. The proposed algorithms are evaluated not only in terms of reconstruction error for different compressions ratios but also in terms of computational performance using two different GPU architectures by NVIDIA: 1) GeForce GTX 590; and 2) GeForce GTX TITAN. Experiments are conducted using both simulated and real data revealing considerable acceleration factors and obtaining good results in the task of compressing remotely sensed hyperspectral data sets.

Fast Spatial Preprocessing for Spectral Unmixing of Hyperspectral Data on Graphics Processing Units

Spectral unmixing is an important technique for hyperspectral data exploitation. It amounts at finding a set of pure spectral signatures (endmembers) of the most representative materials in the scene, and estimating their abundance fractions. The integration of spatial information prior to spectral unmixing of hyperspectral data has attracted much attention in recent years. Several approaches have been developed for the purpose of guiding endmember identification algorithms to spatially representative, yet spectrally pure endmembers. In particular, the spatial preprocessing (SPP) algorithm can be used prior to most existing spectral-based endmember identification techniques, thus promoting the selection of endmembers in spatially representative areas of the scene. However, most SPP techniques are computationally expensive, which adds significant burden to the spectral unmixing process. In this paper, we present three parallel implementations of SPP that have been specifically developed for commodity graphics processing units (GPUs). We perform an evaluation of these techniques using two GPU architectures from NVidia: GeForce GTX 580 and GeForce GTX 870M, which reveals that real-time processing performance can be obtained for real hyperspectral data sets collected by the airborne visible infra-red imaging spectrometer (AVIRIS).

Sparse Unmixing-Based Content Retrieval of Hyperspectral Images on Graphics Processing Units

Content-based image retrieval (CBIR) systems have gained significant importance in the remotely sensed hyperspectral imaging community due to the increasing availability of hyperspectral data collected from different instruments. Spectral unmixing has been a popular technique for not only interpreting hyperspectral images but also retrieving them precisely from databases based on information content. This is due to the fact that the information provided by unmixing (i.e., the spectrally pure components of the scene or endmembers , and their corresponding abundance fractions) provides a very intuitive way to describe the content of the scene in both the spectral and the spatial sense. In this letter, we present a new computationally efficient CBIR system for hyperspectral data (available online: http://hypercomp.es/repositorySparse ) which uses sparse unmixing concepts to retrieve hyperspectral scenes, based on their content, from large repositories. The search is guided by a spectral library, which is used as a guide to retrieve the data in a robust and efficient way. Given the large size of libraries and the sparsity of the unmixing solutions, the incorporation of sparse unmixing to the CBIR engine brings significant advantages. To optimize its performance in computational terms, the system has been implemented in parallel by taking advantage of the computational power of commodity graphics processing units. The proposed system is validated using a collection of synthetic and real hyperspectral images, exhibiting state-of-the-art performance.

Parallel Implementation of Polarimetric Synthetic Aperture Radar Data Processing for Unsupervised Classification Using the Complex Wishart Classifier

This work investigates the parallel implementation of target decomposition and unsupervised classification algorithms for polarimetric synthetic aperture radar (POLSAR) data processing. The algorithms are implemented using two different parallel programming models: 1) clusters of CPUs, using message passing interface (MPI), and 2) commodity graphic processing units (GPUs), using the compute device unified architecture (CUDA). POLSAR data processing generally involves a large amount of computations as the full polarimetric information needs to be decomposed and analyzed. Our experiments reveal that GPU architectures provide a good framework for massive parallelization of POLSAR data processing. For instance, it is found that a single GPU can be more efficient than a cluster of 128 nodes with speedups of more than $100 \times $ in comparison with the single processor times. The proposed implementation makes the best use of low-level features in the GPU architecture such as shared memories, while also providing coalesced accesses to memory in order to achieve maximum performance.