Improving GPGPU Concurrency with Elastic Kernels

Each new generation of GPUs vastly increases the resources available to GPGPU programs. GPU programming models (like CUDA) were designed to scale to use these resources. However, we find that CUDA programs actually do not scale to utilize all available resources, with over 30% of resources going unused on average for programs of the Parboil2 suite that we used in our work. Current GPUs therefore allow concurrent execution of kernels to improve utilization. In this work, we study concurrent execution of GPU kernels using multiprogram workloads on current NVIDIA Fermi GPUs. On two-program workloads from the Parboil2 benchmark suite we find concurrent execution is often no better than serialized execution. We identify that the lack of control over resource allocation to kernels is a major serialization bottleneck. We propose transformations that convert CUDA kernels into elastic kernels which permit fine-grained control over their resource usage. We then propose several elastic-kernel aware concurrency policies that offer significantly better performance and concurrency compared to the current CUDA policy. We evaluate our proposals on real hardware using multiprogrammed workloads constructed from benchmarks in the Parboil 2 suite. On average, our proposals increase system throughput (STP) by 1.21x and improve the average normalized turnaround time (ANTT) by 3.73x for two-program workloads when compared to the current CUDA concurrency implementation.

Técnicas de paralelização em GPGPU aplicadas em algoritmo para remoção de ruído multiplicativo

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES)

Matrix-vector multiplication and triangular linear solver using GPGPU for symmetric positive definite matrices derived from elliptic equations

The modern GPUs are well suited for intensive computational tasks and massive parallel computation. Sparse matrix multiplication and linear triangular solver are the most important and heavily used kernels in scientific computation, and several challenges in developing a high performance kernel with the two modules is investigated. The main interest it to solve linear systems derived from the elliptic equations with triangular elements. The resulting linear system has a symmetric positive definite matrix. The sparse matrix is stored in the compressed sparse row (CSR) format. It is proposed a CUDA algorithm to execute the matrix vector multiplication using directly the CSR format. A dependence tree algorithm is used to determine which variables the linear triangular solver can determine in parallel. To increase the number of the parallel threads, a coloring graph algorithm is implemented to reorder the mesh numbering in a pre-processing phase. The proposed method is compared with parallel and serial available libraries. The results show that the proposed method improves the computation cost of the matrix vector multiplication. The pre-processing associated with the triangular solver needs to be executed just once in the proposed method. The conjugate gradient method was implemented and showed similar convergence rate for all the compared methods. The proposed method showed significant smaller execution time.

General-Purpose Computing on Graphic Processing Unit (GPGPU): metodi e tecnologie per lo sviluppo di applicazioni

Microprocessori basati su singolo processore (CPU), hanno visto una rapida crescita di performances ed un abbattimento dei costi per circa venti anni. Questi microprocessori hanno portato una potenza di calcolo nell’ordine del GFLOPS (Giga Floating Point Operation per Second) sui PC Desktop e centinaia di GFLOPS su clusters di server. Questa ascesa ha portato nuove funzionalità nei programmi, migliori interfacce utente e tanti altri vantaggi. Tuttavia questa crescita ha subito un brusco rallentamento nel 2003 a causa di consumi energetici sempre più elevati e problemi di dissipazione termica, che hanno impedito incrementi di frequenza di clock. I limiti fisici del silicio erano sempre più vicini. Per ovviare al problema i produttori di CPU (Central Processing Unit) hanno iniziato a progettare microprocessori multicore, scelta che ha avuto un impatto notevole sulla comunità degli sviluppatori, abituati a considerare il software come una serie di comandi sequenziali. Quindi i programmi che avevano sempre giovato di miglioramenti di prestazioni ad ogni nuova generazione di CPU, non hanno avuto incrementi di performance, in quanto essendo eseguiti su un solo core, non beneficiavano dell’intera potenza della CPU. Per sfruttare appieno la potenza delle nuove CPU la programmazione concorrente, precedentemente utilizzata solo su sistemi costosi o supercomputers, è diventata una pratica sempre più utilizzata dagli sviluppatori. Allo stesso tempo, l’industria videoludica ha conquistato una fetta di mercato notevole: solo nel 2013 verranno spesi quasi 100 miliardi di dollari fra hardware e software dedicati al gaming. Le software houses impegnate nello sviluppo di videogames, per rendere i loro titoli più accattivanti, puntano su motori grafici sempre più potenti e spesso scarsamente ottimizzati, rendendoli estremamente esosi in termini di performance. Per questo motivo i produttori di GPU (Graphic Processing Unit), specialmente nell’ultimo decennio, hanno dato vita ad una vera e propria rincorsa alle performances che li ha portati ad ottenere dei prodotti con capacità di calcolo vertiginose. Ma al contrario delle CPU che agli inizi del 2000 intrapresero la strada del multicore per continuare a favorire programmi sequenziali, le GPU sono diventate manycore, ovvero con centinaia e centinaia di piccoli cores che eseguono calcoli in parallelo. Questa immensa capacità di calcolo può essere utilizzata in altri campi applicativi? La risposta è si e l’obiettivo di questa tesi è proprio quello di constatare allo stato attuale, in che modo e con quale efficienza pùo un software generico, avvalersi dell’utilizzo della GPU invece della CPU.

Moving object detection strategy for augmented-reality applications in a GPGPU by using CUDA

A spatial-color-based non-parametric background-foreground modeling strategy in a GPGPU by using CUDA is proposed. This strategy is suitable for augmented-reality applications, providing real-time high-quality results in a great variety of scenarios.

Moving object detection for real-time augmented reality applications in a GPGPU

The last generation of consumer electronic devices is endowed with Augmented Reality (AR) tools. These tools require moving object detection strategies, which should be fast and efficient, to carry out higher level object analysis tasks. We propose a lightweight spatio-temporal-based non-parametric background-foreground modeling strategy in a General Purpose Graphics Processing Unit (GPGPU), which provides real-time high-quality results in a great variety of scenarios and is suitable for AR applications.

FTCS finite difference scheme GPGPU parallel computing for the heat conduction equation = Programación en paralelo GPGPU del método en diferencias finitas FTCS para la ecuación del calor

En el presente artículo se muestran las ventajas de la programación en paralelo resolviendo numéricamente la ecuación del calor en dos dimensiones a través del método de diferencias finitas explícito centrado en el espacio FTCS. De las conclusiones de este trabajo se pone de manifiesto la importancia de la programación en paralelo para tratar problemas grandes, en los que se requiere un elevado número de cálculos, para los cuales la programación secuencial resulta impracticable por el elevado tiempo de ejecución. En la primera sección se describe brevemente los conceptos básicos de programación en paralelo. Seguidamente se resume el método de diferencias finitas explícito centrado en el espacio FTCS aplicado a la ecuación parabólica del calor. Seguidamente se describe el problema de condiciones de contorno y valores iniciales específico al que se va a aplicar el método de diferencias finitas FTCS, proporcionando pseudocódigos de una implementación secuencial y dos implementaciones en paralelo. Finalmente tras la discusión de los resultados se presentan algunas conclusiones. In this paper the advantages of parallel computing are shown by solving the heat conduction equation in two dimensions with the forward in time central in space (FTCS) finite difference method. Two different levels of parallelization are consider and compared with traditional serial procedures. We show in this work the importance of parallel computing when dealing with large problems that are impractical or impossible to solve them with a serial computing procedure. In the first section a summary of parallel computing approach is presented. Subsequently, the forward in time central in space (FTCS) finite difference method for the heat conduction equation is outline, describing how the heat flow equation is derived in two dimensions and the particularities of the finite difference numerical technique considered. Then, a specific initial boundary value problem is solved by the FTCS finite difference method and serial and parallel pseudo codes are provided. Finally after results are discussed some conclusions are presented.

GPGPU implementation of growing neural gas: application to 3D scene reconstruction

Self-organising neural models have the ability to provide a good representation of the input space. In particular the Growing Neural Gas (GNG) is a suitable model because of its flexibility, rapid adaptation and excellent quality of representation. However, this type of learning is time-consuming, especially for high-dimensional input data. Since real applications often work under time constraints, it is necessary to adapt the learning process in order to complete it in a predefined time. This paper proposes a Graphics Processing Unit (GPU) parallel implementation of the GNG with Compute Unified Device Architecture (CUDA). In contrast to existing algorithms, the proposed GPU implementation allows the acceleration of the learning process keeping a good quality of representation. Comparative experiments using iterative, parallel and hybrid implementations are carried out to demonstrate the effectiveness of CUDA implementation. The results show that GNG learning with the proposed implementation achieves a speed-up of 6× compared with the single-threaded CPU implementation. GPU implementation has also been applied to a real application with time constraints: acceleration of 3D scene reconstruction for egomotion, in order to validate the proposal.

Real-time 3D semi-local surface patch extraction using GPGPU

Feature vectors can be anything from simple surface normals to more complex feature descriptors. Feature extraction is important to solve various computer vision problems: e.g. registration, object recognition and scene understanding. Most of these techniques cannot be computed online due to their complexity and the context where they are applied. Therefore, computing these features in real-time for many points in the scene is impossible. In this work, a hardware-based implementation of 3D feature extraction and 3D object recognition is proposed to accelerate these methods and therefore the entire pipeline of RGBD based computer vision systems where such features are typically used. The use of a GPU as a general purpose processor can achieve considerable speed-ups compared with a CPU implementation. In this work, advantageous results are obtained using the GPU to accelerate the computation of a 3D descriptor based on the calculation of 3D semi-local surface patches of partial views. This allows descriptor computation at several points of a scene in real-time. Benefits of the accelerated descriptor have been demonstrated in object recognition tasks. Source code will be made publicly available as contribution to the Open Source Point Cloud Library.

Towards a real-time 3D object recognition pipeline on mobile GPGPU computing platforms using low-cost RGB-D sensors

In this project, we propose the implementation of a 3D object recognition system which will be optimized to operate under demanding time constraints. The system must be robust so that objects can be recognized properly in poor light conditions and cluttered scenes with significant levels of occlusion. An important requirement must be met: the system must exhibit a reasonable performance running on a low power consumption mobile GPU computing platform (NVIDIA Jetson TK1) so that it can be integrated in mobile robotics systems, ambient intelligence or ambient assisted living applications. The acquisition system is based on the use of color and depth (RGB-D) data streams provided by low-cost 3D sensors like Microsoft Kinect or PrimeSense Carmine. The range of algorithms and applications to be implemented and integrated will be quite broad, ranging from the acquisition, outlier removal or filtering of the input data and the segmentation or characterization of regions of interest in the scene to the very object recognition and pose estimation. Furthermore, in order to validate the proposed system, we will create a 3D object dataset. It will be composed by a set of 3D models, reconstructed from common household objects, as well as a handful of test scenes in which those objects appear. The scenes will be characterized by different levels of occlusion, diverse distances from the elements to the sensor and variations on the pose of the target objects. The creation of this dataset implies the additional development of 3D data acquisition and 3D object reconstruction applications. The resulting system has many possible applications, ranging from mobile robot navigation and semantic scene labeling to human-computer interaction (HCI) systems based on visual information.

Aceleración de Algoritmos en Procesadores Multicore, GPGPU y FPGA

El avance en la potencia de cómputo en nuestros días viene dado por la paralelización del procesamiento, dadas las características que disponen las nuevas arquitecturas de hardware. Utilizar convenientemente este hardware impacta en la aceleración de los algoritmos en ejecución (programas). Sin embargo, convertir de forma adecuada el algoritmo en su forma paralela es complejo, y a su vez, esta forma, es específica para cada tipo de hardware paralelo. En la actualidad los procesadores de uso general más comunes son los multicore, procesadores paralelos, también denominados Symmetric Multi-Processors (SMP). Hoy en día es difícil hallar un procesador para computadoras de escritorio que no tengan algún tipo de paralelismo del caracterizado por los SMP, siendo la tendencia de desarrollo, que cada día nos encontremos con procesadores con mayor numero de cores disponibles. Por otro lado, los dispositivos de procesamiento de video (Graphics Processor Units - GPU), a su vez, han ido desarrollando su potencia de cómputo por medio de disponer de múltiples unidades de procesamiento dentro de su composición electrónica, a tal punto que en la actualidad no es difícil encontrar placas de GPU con capacidad de 200 a 400 hilos de procesamiento paralelo. Estos procesadores son muy veloces y específicos para la tarea que fueron desarrollados, principalmente el procesamiento de video. Sin embargo, como este tipo de procesadores tiene muchos puntos en común con el procesamiento científico, estos dispositivos han ido reorientándose con el nombre de General Processing Graphics Processor Unit (GPGPU). A diferencia de los procesadores SMP señalados anteriormente, las GPGPU no son de propósito general y tienen sus complicaciones para uso general debido al límite en la cantidad de memoria que cada placa puede disponer y al tipo de procesamiento paralelo que debe realizar para poder ser productiva su utilización. Los dispositivos de lógica programable, FPGA, son dispositivos capaces de realizar grandes cantidades de operaciones en paralelo, por lo que pueden ser usados para la implementación de algoritmos específicos, aprovechando el paralelismo que estas ofrecen. Su inconveniente viene derivado de la complejidad para la programación y el testing del algoritmo instanciado en el dispositivo. Ante esta diversidad de procesadores paralelos, el objetivo de nuestro trabajo está enfocado en analizar las características especificas que cada uno de estos tienen, y su impacto en la estructura de los algoritmos para que su utilización pueda obtener rendimientos de procesamiento acordes al número de recursos utilizados y combinarlos de forma tal que su complementación sea benéfica. Específicamente, partiendo desde las características del hardware, determinar las propiedades que el algoritmo paralelo debe tener para poder ser acelerado. Las características de los algoritmos paralelos determinará a su vez cuál de estos nuevos tipos de hardware son los mas adecuados para su instanciación. En particular serán tenidos en cuenta el nivel de dependencia de datos, la necesidad de realizar sincronizaciones durante el procesamiento paralelo, el tamaño de datos a procesar y la complejidad de la programación paralela en cada tipo de hardware.

The Queensland Cancer Physics Collaborative and the Quantitative Biomedical Imaging & Characterisation (Q-BIC) Research Group

a presentation about immersive visualised simulation systems, image analysis and GPGPU Techonology

On self-interacting oligoribonucleotides. I. Absorption and optical rotatory dispersion of 2′-5′- and 3′-5′-oligoguanylic acids

A series of 2′-5′-oligoguanylic acids are prepared by reacting G(cyclic)p with takadiastase T1 ribonuclease and separating the products chromatographically. The 3′-5′-oligoguanylic acids are obtained by separating the products of alkaline degradation of 3′-5′-poly(G). The optical rotatory dispersion and hypochromism of both 2′-5′- and 3′-5′-oligoguanylic acids are studied at two different pH. The optical rotatory dispersion spectrum of 2′-5′-GpG is significantly different from that of 3′-5′-GpG. The magnitude of rotation of the long-wavelength peak of 2′-5′-GpG is larger than that of 3′-5′-GpG. This finding contradicts the explanation that the extra stability and more intense circular dichroism band of other 3′-5′-dinucleoside monophosphates is due to H-bond formation between 2′-OH and either the base or the phosphate oxygen. The end phosphate group has a marked effect on the spectrum of GpG between 230 and 250 mμ. In addition the optical rotatory dispersion spectra of 2′-5′ exhibit strong pH, temperature, and solvent dependence between 230 and 250 mμ. ΔH and AS for order ⇌ disorder transition is estimated to be 9.7 kcal/mole and 35.2 eu, respectively. The optical rotatory dispersion spectra of guanine-rich oligoribonucleotides, GpGpC, GpGpU, GpGpGpC, and GpGpGpU are compared to the calculated optical rotatory dispersion from the semiempirical expression of Cantor and Tinoco, using measured optical rotatory dispersion of dimers. Contrary to previous studies, agreement is found not at all satisfactory. However, optical rotatory dispersion of 3′-5′-GpGpGpC and GpGpGpU can be estimated from the semiempirical expression, if a next-nearest interaction parameter is introduced empirically. Such interaction parameter can be calculated from the measured properties of trinucleotide sequences like GpGpG, GpGpC, and GpGpU, assuming that only the nearest-neighbor interaction is important. The optical rotatory dispersion of single-stranded poly(G) is also predicted. The importance of syn-anti equilibrium and next-nearest-neighbor interaction in oligoguanylic acids is suggested as a probable explanation.

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.

Taming Control Divergence in GPUs through Control Flow Linearization

Branch divergence is a very commonly occurring performance problem in GPGPU in which the execution of diverging branches is serialized to execute only one control flow path at a time. Existing hardware mechanism to reconverge threads using a stack causes duplicate execution of code for unstructured control flow graphs. Also the stack mechanism cannot effectively utilize the available parallelism among diverging branches. Further, the amount of nested divergence allowed is also limited by depth of the branch divergence stack. In this paper we propose a simple and elegant transformation to handle all of the above mentioned problems. The transformation converts an unstructured CFG to a structured CFG without duplicating user code. It incurs only a linear increase in the number of basic blocks and also the number of instructions. Our solution linearizes the CFG using a predicate variable. This mechanism reconverges the divergent threads as early as possible. It also reduces the depth of the reconvergence stack. The available parallelism in nested branches can be effectively extracted by scheduling the basic blocks to reduce the effect of stalls due to memory accesses. It can also increase execution efficiency of nested loops with different trip counts for different threads. We implemented the proposed transformation at PTX level using the Ocelot compiler infrastructure. We evaluated the technique using various benchmarks to show that it can be effective in handling the performance problem due to divergence in unstructured CFGs.