Ever since chip manufacturers hit the power wall preventing them from increasing processor clock speed, there has been an increased push towards parallelism for performance improvements. This parallelism comes in the form of both data parallel single instruction multiple data (SIMD) instructions, as well as parallel compute cores in both central processing units (CPUs) and graphics processing units (GPUs). While these hardware enhancements offer potential performance enhancements, programs must be re-written to take advantage of them in order to see any performance improvement Some lower level languages that compile directly to machine code already take advantage of the data parallel SIMD instructions, but often higher level interpreted languages do not. Java, one of the most popular programming languages in the world, still does not include support for these SIMD instructions. In this thesis, we present a vector library that implements all of the major SIMD instructions in functions that are accessible to Java through JNI function calls. This brings the benefits of general purpose SIMD functionality to Java. This thesis also works with the data parallel Aparapi Java extension to bring these SIMD performance improvements to programmers who use the extension without any additional effort on their part. Aparapi already provides programmers with an API that allows programmers to declare certain sections of their code parallel. These parallel sections are then run on OpenCL capable hardware with a fallback path in the Java thread pool to ensure code reliability. This work takes advantage of the knowledge of independence of the parallel sections of code to automatically modify the Java thread pool fallback path to include the vectorization library through the use of an auto-vectorization tool created for this work. When the code is not vectorizable the auto-vectorizer tool is still able to offer performance improvements over the default fallback path through an improved looped implementation that executes the same code but with less overhead. Experiments conducted by this work illustrate that for all 10 benchmarks tested the auto-vectorization tool was able to produce an implementation that was able to beat the default Aparapi fallback path. In addition it was found that this improved fallback path even outperformed the GPU implementation for several of the benchmarks tested. / Master of Science

Cognitive radio is a growing field in communications which allows a radio to automatically configure its transmission or reception properties in order to reduce interference, provide better quality of service, or allow for more users in a given spectrum. Such processes require several complex features that are currently being utilized in cognitive radio. Two such features, spectrum sensing and identification, have been implemented in numerous ways, however, they generally suffer from high computational complexity. Additionally, Hidden Markov Models (HMMs) are a widely used mathematical modeling tool used in various fields of engineering and sciences. In electrical and computer engineering, it is used in several areas, including speech recognition, handwriting recognition, artificial intelligence, queuing theory, and are used to model fading in communication channels. The research presented in this thesis proposes a new approach to spectrum identification using a parallel implementation of Hidden Markov Models. Algorithms involving HMMs are usually implemented in the traditional serial manner, which have prohibitively long runtimes. In this work, we study their use in parallel implementations and compare our approach to traditional serial implementations. Timing and power measurements are taken and used to show that the parallel implementation can achieve well over 100Ã speedup in certain situations. To demonstrate the utility of this new parallel algorithm using graphics processing units (GPUs), a new method for signal identification is proposed for both serial and parallel implementations using HMMs. The method achieved high recognition at -10 dB Eb/N0. HMMs can benefit from parallel implementation in certain circumstances, specifically, in models that have many states or when multiple models are used in conjunction. / Master of Science

The use of hardware accelerators in high-performance computing has grown increasingly prevalent, particularly due to the growth of graphics processing units (GPUs) as general-purpose (GPGPU) accelerators. Much of this growth has been driven by NVIDIA's CUDA ecosystem for developing GPGPU applications on NVIDIA hardware. However, with the increasing diversity of GPUs (including those from AMD, ARM, and Qualcomm), OpenCL has emerged as an open and vendor-agnostic environment for programming GPUs as well as other parallel computing devices such as the CPU (central processing unit), APU (accelerated processing unit), FPGA (field programmable gate array), and DSP (digital signal processor). The above, coupled with the broader array of devices supporting OpenCL and the significant conceptual and syntactic overlap between CUDA and OpenCL, motivated the creation of a CUDA-to-OpenCL source-to-source translator. However, there exist sufficient differences that make the translation non-trivial, providing practical limitations to both manual and automatic translation efforts. In this thesis, the performance, coverage, and reliability of a prototype CUDA-to-OpenCL source translator are addressed via extensive profiling of a large body of sample CUDA applications. An analysis of the sample body of applications is provided, which identifies and characterizes general CUDA source constructs and programming practices that obstruct our translation efforts. This characterization then led to more robust support for the translator, followed by an evaluation that demonstrated the performance of our automatically-translated OpenCL is on par with the original CUDA for a subset of sample applications when executed on the same NVIDIA device. / Master of Science

Graphics Processing Units (GPUs) are workhorses of modern performance due to their ability to achieve massive speedups on parallel applications. The massive number of threads that can be run concurrently on these systems allow applications which have data-parallel computations to achieve better performance when compared to traditional CPU systems. However, the GPU is not perfect for all types of computation. The massively parallel SIMT architecture of the GPU can still be constraining in terms of achievable performance. GPU-based systems will typically only be able to achieve between 40%-60% of their peak performance. One of the major problems affecting this effeciency is the GPU memory system, which is tailored to the needs of graphics workloads instead of general-purpose computation. This thesis intends to show the importance of memory optimizations for GPU systems. In particular, this work addresses problems of data transfer and global atomic memory contention. Using the novel AMD Fusion architecture, we gain overall performance improvements over discrete GPU systems for data-intensive applications. The fused architecture systems offer an interesting trade off by increasing data transfer rates at the cost of some raw computational power. We characterize the performance of different memory paths that are possible because of the shared memory space present on the fused architecture. In addition, we provide a theoretical model which can be used to correctly predict the comparative performance of memory movement techniques for a given data-intensive application and system. In terms of global atomic memory contention, we show improvements in scalability and performance for global synchronization primitives by avoiding contentious global atomic memory accesses. In general, this work shows the importance of understanding the memory system of the GPU architecture to achieve better application performance. / Master of Science

Graphics Processing Units (GPUs) are being adopted in many computing facilities given their extraordinary computing power, which makes it possible to accelerate many general purpose applications from different domains. However, GPUs also present several side effects, such as increased acquisition costs as well as larger space requirements. They also require more powerful energy supplies. Furthermore, GPUs still consume some amount of energy while idle and their utilization is usually low for most workloads. In a similar way to virtual machines, the use of virtual GPUs may address the aforementioned concerns. In this regard, the remote GPU virtualization mechanism allows an application being executed in a node of the cluster to transparently use the GPUs installed at other nodes. Moreover, this technique allows to share the GPUs present in the computing facility among the applications being executed in the cluster. In this way, several applications being executed in different (or the same) cluster nodes can share one or more GPUs located in other nodes of the cluster. Sharing GPUs should increase overall GPU utilization, thus reducing the negative impact of the side effects mentioned before. Reducing the total amount of GPUs installed in the cluster may also be possible. In this dissertation we enhance one framework offering remote GPU virtualization capabilities, referred to as rCUDA, for its use in high-performance clusters. While the initial prototype version of rCUDA demonstrated its functionality, it also revealed concerns with respect to usability, performance, and support for new GPU features, which prevented its used in production environments. These issues motivated this thesis, in which all the research is primarily conducted with the aim of turning rCUDA into a production-ready solution for eventually transferring it to industry. The new version of rCUDA resulting from this work presents a reduction of up to 35% in execution time of the applications analyzed with respect to the initial version. Compared to the use of local GPUs, the overhead of this new version of rCUDA is below 5% for the applications studied when using the latest high-performance computing networks available. / Las unidades de procesamiento gráfico (Graphics Processing Units, GPUs) están siendo utilizadas en muchas instalaciones de computación dada su extraordinaria capacidad de cálculo, la cual hace posible acelerar muchas aplicaciones de propósito general de diferentes dominios. Sin embargo, las GPUs también presentan algunas desventajas, como el aumento de los costos de adquisición, así como mayores requerimientos de espacio. Asimismo, también requieren un suministro de energía más potente. Además, las GPUs consumen una cierta cantidad de energía aún estando inactivas, y su utilización suele ser baja para la mayoría de las cargas de trabajo. De manera similar a las máquinas virtuales, el uso de GPUs virtuales podría hacer frente a los inconvenientes mencionados. En este sentido, el mecanismo de virtualización remota de GPUs permite que una aplicación que se ejecuta en un nodo de un clúster utilice de forma transparente las GPUs instaladas en otros nodos de dicho clúster. Además, esta técnica permite compartir las GPUs presentes en el clúster entre las aplicaciones que se ejecutan en el mismo. De esta manera, varias aplicaciones que se ejecutan en diferentes nodos de clúster (o los mismos) pueden compartir una o más GPUs ubicadas en otros nodos del clúster. Compartir GPUs aumenta la utilización general de la GPU, reduciendo así el impacto negativo de las desventajas anteriormente mencionadas. De igual forma, este mecanismo también permite reducir la cantidad total de GPUs instaladas en el clúster. En esta tesis mejoramos un entorno de trabajo llamado rCUDA, el cual ofrece funcionalidades de virtualización remota de GPUs para su uso en clusters de altas prestaciones. Si bien la versión inicial del prototipo de rCUDA demostró su funcionalidad, también reveló dificultades con respecto a la usabilidad, el rendimiento y el soporte para nuevas características de las GPUs, lo cual impedía su uso en entornos de producción. Estas consideraciones motivaron la presente tesis, en la que toda la investigación llevada a cabo tiene como objetivo principal convertir rCUDA en una solución lista para su uso entornos de producción, con la finalidad de transferirla eventualmente a la industria. La nueva versión de rCUDA resultante de este trabajo presenta una reducción de hasta el 35% en el tiempo de ejecución de las aplicaciones analizadas con respecto a la versión inicial. En comparación con el uso de GPUs locales, la sobrecarga de esta nueva versión de rCUDA es inferior al 5% para las aplicaciones estudiadas cuando se utilizan las últimas redes de computación de altas prestaciones disponibles. / Les unitats de processament gràfic (Graphics Processing Units, GPUs) estan sent utilitzades en moltes instal·lacions de computació donada la seva extraordinària capacitat de càlcul, la qual fa possible accelerar moltes aplicacions de propòsit general de diferents dominis. No obstant això, les GPUs també presenten alguns desavantatges, com l'augment dels costos d'adquisició, així com major requeriment d'espai. Així mateix, també requereixen un subministrament d'energia més potent. A més, les GPUs consumeixen una certa quantitat d'energia encara estant inactives, i la seua utilització sol ser baixa per a la majoria de les càrregues de treball. D'una manera semblant a les màquines virtuals, l'ús de GPUs virtuals podria fer front als inconvenients esmentats. En aquest sentit, el mecanisme de virtualització remota de GPUs permet que una aplicació que s'executa en un node d'un clúster utilitze de forma transparent les GPUs instal·lades en altres nodes d'aquest clúster. A més, aquesta tècnica permet compartir les GPUs presents al clúster entre les aplicacions que s'executen en el mateix. D'aquesta manera, diverses aplicacions que s'executen en diferents nodes de clúster (o els mateixos) poden compartir una o més GPUs ubicades en altres nodes del clúster. Compartir GPUs augmenta la utilització general de la GPU, reduint així l'impacte negatiu dels desavantatges anteriorment esmentades. A més a més, aquest mecanisme també permet reduir la quantitat total de GPUs instal·lades al clúster. En aquesta tesi millorem un entorn de treball anomenat rCUDA, el qual ofereix funcionalitats de virtualització remota de GPUs per al seu ús en clústers d'altes prestacions. Si bé la versió inicial del prototip de rCUDA va demostrar la seua funcionalitat, també va revelar dificultats pel que fa a la usabilitat, el rendiment i el suport per a noves característiques de les GPUs, la qual cosa impedia el seu ús en entorns de producció. Aquestes consideracions van motivar la present tesi, en què tota la investigació duta a terme té com a objectiu principal convertir rCUDA en una solució preparada per al seu ús entorns de producció, amb la finalitat de transferir-la eventualment a la indústria. La nova versió de rCUDA resultant d'aquest treball presenta una reducció de fins al 35% en el temps d'execució de les aplicacions analitzades respecte a la versió inicial. En comparació amb l'ús de GPUs locals, la sobrecàrrega d'aquesta nova versió de rCUDA és inferior al 5% per a les aplicacions estudiades quan s'utilitzen les últimes xarxes de computació d'altes prestacions disponibles. / Reaño González, C. (2017). On the Enhancement of Remote GPU Virtualization in High Performance Clusters [Tesis doctoral]. Universitat Politècnica de València. https://doi.org/10.4995/Thesis/10251/86219 / Premios Extraordinarios de tesis doctorales

Durant cette thèse, nous avons principalement travaillé sur deux sujets liés à l'analyse de la performance GPU (Graphics Processing Unit - Processeur graphique). Dans un premier temps, nous avons développé une méthode analytique et un outil d'estimation temporel (TEG) pour prédire les performances d'applications CUDA s’exécutant sur des GPUs de la famille GT200. Cet outil peut prédire les performances avec une précision approchant celle des outils précis au cycle près. Dans un second temps, nous avons développé une approche pour estimer la borne supérieure des performances d'une application GPU, en se basant sur l'analyse de l'application et de son code assembleur. Avec cette borne, nous connaissons la marge d'optimisation restante, et nous pouvons décider des efforts d'optimisation à fournir. Grâce à cette analyse, nous pouvons aussi comprendre quels paramètres sont critiques à la performance. / In this thesis work, we have mainly worked on two topics of GPU performance analysis. First, we have developed an analytical method and a timing estimation tool (TEG) to predict CUDA application's performance for GT200 generation GPUs. TEG can predict GPU applications' performance in cycle-approximate level. Second, we have developed an approach to estimate GPU applications' performance upper bound based on application analysis and assembly code level benchmarking. With the performance upper bound of an application, we know how much optimization space is left and can decide the optimization effort. Also with the analysis we can understand which parameters are critical to the performance.

La donnée sismique depuis sa conception (modélisation d’acquisitions sismiques), dans sa phase de traitement (prétraitement et migration) et jusqu’à son exploitation pour en extraire les informations géologiques pertinentes nécessaires à l’identification et l’exploitation optimale des réservoirs d’hydrocarbures (interprétation), génère un volume important de calculs. Nous montrons dans ce travail de thèse qu’à chacune de ces étapes l’utilisation de technologies accélératrices de type GPGPU permet de réduire radicalement les temps de calcul tout en restant dans une enveloppe de consommation électrique raisonnable. Nous présentons et analysons les éléments sous-jacents à ces performances. L’importance de l’utilisation de motifs d’accès mémoire adéquats est particulièrement mise en exergue étant donné que l’accès à la mémoire représente le principal goulot d’étranglement pour les algorithmes abordés. Nous reportons des facteurs d’accélération de l’ordre de 40 pour la modélisation sismique par résolution de l’équation d’onde par différences finies (brique de base pour la modélisation et l’imagerie sismique) et entre 8 et 113 pour le calcul d’attributs sismiques. Nous démontrons que l’utilisation d’accélérateurs matériels élargit considérablement le champ du possible, aussi bien en imagerie sismique (modélisation de nouveaux types d’acquisitions à grande échelle) qu’en interprétation (calcul d’attributs complexes sur station de travail, paramétrage interactif des calculs, etc.). / During the seismic imaging workflow, from seismic modeling to interpretation, processingseismic data requires a massive amount of computation. We show in this work that, at eachstage of this workflow, hardware accelerators such as GPUs may help reducing the time requiredto process seismic data while staying at reasonable energy consumption levels.In this work, the key programming considerations needed to achieve good performance are describedand discussed. The importance of adapted in-memory data access patterns is particularlyemphasised since data access is the main bottleneck for the considered algorithms. When usingGPUs, speedup ratios of 40× are achieved for FDTD seismic modeling, and 8× up to 113× forseismic attribute computation compared to CPUs.

[pt] Sistemas de equações lineares esparsos e de grande porte aparecem como resultado da modelagem de vários problemas nas engenharias. Dada sua importância, muitos trabalhos estudam métodos para a resolução desses sistemas. Esta dissertação explora o potencial computacional de múltiplas GPUs, utilizando a tecnologia OpenCL, com a finalidade de resolver sistemas de equações lineares de grande porte. Na metodologia proposta, o método do gradiente conjugado é subdivido em kernels que são resolvidos por múltiplas GPUs. Para tal, se fez necessário compreender como a arquitetura das GPUs se relaciona com a tecnologia OpenCL a fim de obter um melhor desempenho. / [en] The process of modeling problems in the engineering fields tends to produce substantiously large systems of sparse linear equations. Extensive research has been done to devise methods to solve these systems. This thesis explores the computational potential of multiple GPUs, through the use of the OpenCL tecnology, aiming to tackle the solution of large systems of sparse linear equations. In the proposed methodology, the conjugate gradient method is subdivided into kernels, which are delegated to multiple GPUs. In order to achieve an efficient method, it was necessary to understand how the GPUs’ architecture communicates with OpenCL.

Un robot humanoïde est un système complexe doté de nombreux degrés de liberté, et dont le comportement est sujet aux équations non linéaires du mouvement. Par conséquent, la planification de mouvement pour un tel système est une tâche difficile d'un point de vue calculatoire. Dans ce mémoire, nous avons pour objectif de développer une méthode permettant d'utiliser la puissance de calcul des GPUs dans le contexte de la planification de mouvement corps-complet basée sur de l'optimisation. Nous montrons dans un premier temps les propriétés du problème d'optimisation, et des pistes d'étude pour la parallélisation de ce dernier. Ensuite, nous présentons notre approche du calcul de la dynamique, adaptée aux architectures de calcul parallèle. Cela nous permet de proposer une implémentation de notre problème de planification de mouvement sur GPU: contraintes et gradients sont calculés en parallèle, tandis que la résolution du problème même se déroule sur le CPU. Nous proposons en outre une nouvelle paramétrisation des forces de contact adaptée à notre problème d'optimisation. Enfin, nous étudions l'extension de notre travail au contrôle prédictif. / A humanoid robot is a complex system with numerous degrees of freedom, whose behavior is subject to the nonlinear equations of motion. As a result, planning its motion is a difficult task from a computational perspective.In this thesis, we aim at developing a method that can leverage the computing power of GPUs in the context of optimization-based whole-body motion planning. We first exhibit the properties of the optimization problem, and show that several avenues can be exploited in the context of parallel computing. Then, we present our approach of the dynamics computation, suitable for highly-parallel processing architectures. Next, we propose a many-core GPU implementation of the motion planning problem. Our approach computes the constraints and their gradients in parallel, and feeds the result to a nonlinear optimization solver running on the CPU. Because each constraint and its gradient can be evaluated independently for each time interval, we end up with a highly parallelizable problem that can take advantage of GPUs. We also propose a new parametrization of contact forces adapted to our optimization problem. Finally, we investigate the extension of our work to model predictive control.

Este trabalho apresenta os resultados da pesquisa e da aplicação de técnicas de GPGPU (General-Purpose computation on Graphics Processing Units) sobre o sistema de vídeo-avatar com realidade aumentada denominado AVMix. Com o aumento da demanda por gráficos tridimensionais interativos em tempo real cada vez mais próximos da realidade, as GPUs (Graphics Processing Units) evoluíram até o estado atual, como um hardware com alto poder computacional que permite o processamento de algoritmos paralelamente sobre um grande volume de dados. Desta forma, É possível usar esta capacidade para aumentar o desempenho de algoritmos usados em diversas áreas, tais como a área de processamento de imagens e visão computacional. A partir das pesquisas de trabalhos semelhantes, definiu-se o uso da arquitetura CUDA (Computer Unified Device Architecture) da Nvidia, que facilita a implementação dos programas executados na GPU e ao mesmo tempo flexibiliza o seu uso, expondo ao programador o detalhamento de alguns recursos de hardware, como por exemplo a quantidade de processadores alocados e os diferentes tipos de memória. Após a reimplementação das rotinas críticas ao desempenho do sistema AVMix (mapa de profundidade, segmentação e interação), os resultados mostram viabilidade do uso da GPU para o processamento de algoritmos paralelos e a importância da avaliação do algoritmo a ser implementado em relação a complexidade do cálculo e ao volume de dados transferidos entre a GPU e a memória principal do computador. / This work presents the results of research and application of GPGPU (General-Purpose computation on Graphics Processing Units) techniques on the video-avatar system with augmented reality called AVMix. With increasing demand for interactive three-dimensional graphics rendered in real-time and closer to reality, GPUs (Graphics Processing Units) evolved to the present state as a high-powered computing hardware enabled to process parallel algorithms over a large data set. This way, it is possible to use this capability to increase the performance of algorithms used in several areas, such as image processing and computer vision. From the research of similar work, it is possible to define the use of CUDA (Computer Unified Device Architecture) from Nvidia, which facilitates the implementation of the programs that run on GPU and at the same time flexibilize its use, exposing to the programmer some details of hardware such as the number of processors allocated and the different types of memory. Following the reimplementation of critical performance routines of AVMix system (depth map, segmentation and interaction), the results show the viability of using the GPU to process parallel algorithms in this application and the importance of evaluating the algorithm to be implemented, considering the complexity of the calculation and the volume of data transferred between the GPU and the computer\'s main memory.