Mathematicians and computational scientists are often limited in their ability to model complex phenomena by the time it takes to run simulations. This thesis will inform interested researchers on how the development of highly parallel computer graphics hardware and the compiler frameworks to exploit it are expanding the range of algorithms that can be explored on affordable commodity hardware. We will discuss the complexities that have prevented researchers from exploiting advanced hardware as well as the obstacles that remain for the non-computer scientist.

Uždengtos geometrijos atrinkimas – tai būdas nustatyti geometriją, kuri yra uždengta su kita geometrija ir dėl to gali būti nevaizduojama, nes neturės jokios įtakos vaizduojamam paveikslui. Tokios geometrijos nevaizdavimas didina vaizdavimo procedūros našumą. Egzistuoja eilė uždengtos geometrijos nustatymo būdų, iš kurių vienas yra hierarchinis uždengtos geometrijos atrinkimo algoritmas. Šiame darbe yra analizuojami uždengtos geometrijos nustatymo būdai bei nagrinėjamos pasirinkto algoritmo veikimo spartinimo galimybės panaudojus DirectCompute technologiją. Ši technologija yra Microsoft DirectX 11 bibliotekų rinkinio dalis, kuri leidžia panaudoti grafinį procesorių bendro pobūdžio skaičiavimams. Darbe iškeltų tikslų pasiekimui yra realizuotos kelios bazinės algoritmo versijos modifikacijos, atliekami modifikuotų versijų veikimo laiko bei įvairių veikimo laiką įtakojančių faktorių tyrimai. Yra aptariami gauti rezultatai bei pateikiamos išvados. / Occlusion culling is a method, which task is to determine geometry occluded with other geometry. Rendering this geometry is useless because it wouldn’t impact rendered picture in any way, so discarding it will improve render time. There are various methods to determine occluded geometry and hierarchical occlusion culling is one of them. This document contains a short summary of these methods, but it’s mainly focused on improving hierarchical occlusion culling algorithm performance by making use of DirectCompute technology. This technology is a part of Microsoft DirectX 11 API, which helps the developer to use graphics processor for general-purpose computation. Main goal is reached by performing in-depth analysis of implemented hierarchical occlusion culling algorithm modifications. This analysis consists of both general performance and various performance-related analyses. Further down the road conclusions and recommendations are given based on performed work and overall results.

Este trabalho descreve o desenvolvimento de um sistema de \"software\" para processamento, visualização e manipulação de imagens tomográficas obtidas por RM. Primeiramente, foi desenvolvido um sistema para uma estação de trabalho Interpro 32&#176C, da Intergraph. Este sistema foi desenvolvido de forma amigável, utilizando-se janelas e \"menus\", para efetuar a interação com o usuário. Foram analisadas e implementadas várias operações de processamento de imagens, tais como: filtros digitais, manipulação da imagem através do seu histograma e medidas estatísticas sobre regiões da imagem. A experiência e o conhecimento obtido com esta implementação, (oram utilizados para definir e desenvolver um novo ambiente gráfico, para uma arquitetura de \"hardware\" baseada no processador gráfico TMS34010 da Texas Instruments, previamente desenvolvida no Laboratório de Instrumentação Eletrônica (LIE). Este ambiente, que suporta menus e janelas, foi desenvolvido utilizando-se das instruções gráficas inerentes à família TMS340. Foram desenvolvidas rotinas de comunicação com o hospedeiro, as quais permitem, por exemplo, que as imagens armazenadas no hospedeiro sejam transmitidas para a arquitetura dedicada. As mesmas rotinas foram utilizadas para depurar e validar o sistema gráfico. / The present work describes the development of a software system to process, visualize and manipulate Magnetic Resonance Images. First an application software W8S developed for an Interpro 32&#176C Intergraph workstation. In this menu and window based 8Oftware, image operations such as filters, transformations and image manipulation based on histograms with statistical data was implemented and evaluated. The knowledge and experience obtained with this implementation was subsequently used to define and design a new graphic environment for a dedicated TMS340 based hardware architecture previously developed in our laboratory. This new toolkit provides window and menu support, and was developed through extensive use of the powerful TMS340 native graphics instructions. Debugging and tracing routines were developed based on an implemented host (PC) - graphics processor board communication system. The same communication software was also used in the examples to perform menu selected downloads of images and functions from the host.

<p>In this report, we show how to implement direct ray tracing of B´ezier surfaces on graphics processing units (GPUs), in particular bicubic rectangular Bézier surfaces and nonparametric cubic Bézier triangles. We use Newton’s method for the rectangular case and show how to use this method to find the ray-surface intersection. For Newton’s method to work we must build a spatial partitioning hierarchy around each surface patch, and in general, hierarchies are essential to speed up the process of ray tracing. We have chosen to use bounding box hierarchies and show how to implement stackless traversal of such a structure on a GPU. For the nonparametric triangular case, we show how to find the wanted intersection by simply solving a cubic polynomial. Because of the limited precision of current GPUs, we also propose a numerical approach to solve the problem, using a one-dimensional Newton search.</p>

In this report, we show how to implement direct ray tracing of B´ezier surfaces on graphics processing units (GPUs), in particular bicubic rectangular Bézier surfaces and nonparametric cubic Bézier triangles. We use Newton’s method for the rectangular case and show how to use this method to find the ray-surface intersection. For Newton’s method to work we must build a spatial partitioning hierarchy around each surface patch, and in general, hierarchies are essential to speed up the process of ray tracing. We have chosen to use bounding box hierarchies and show how to implement stackless traversal of such a structure on a GPU. For the nonparametric triangular case, we show how to find the wanted intersection by simply solving a cubic polynomial. Because of the limited precision of current GPUs, we also propose a numerical approach to solve the problem, using a one-dimensional Newton search.

Este trabalho descreve o desenvolvimento de um sistema de \"software\" para processamento, visualização e manipulação de imagens tomográficas obtidas por RM. Primeiramente, foi desenvolvido um sistema para uma estação de trabalho Interpro 32&#176C, da Intergraph. Este sistema foi desenvolvido de forma amigável, utilizando-se janelas e \"menus\", para efetuar a interação com o usuário. Foram analisadas e implementadas várias operações de processamento de imagens, tais como: filtros digitais, manipulação da imagem através do seu histograma e medidas estatísticas sobre regiões da imagem. A experiência e o conhecimento obtido com esta implementação, (oram utilizados para definir e desenvolver um novo ambiente gráfico, para uma arquitetura de \"hardware\" baseada no processador gráfico TMS34010 da Texas Instruments, previamente desenvolvida no Laboratório de Instrumentação Eletrônica (LIE). Este ambiente, que suporta menus e janelas, foi desenvolvido utilizando-se das instruções gráficas inerentes à família TMS340. Foram desenvolvidas rotinas de comunicação com o hospedeiro, as quais permitem, por exemplo, que as imagens armazenadas no hospedeiro sejam transmitidas para a arquitetura dedicada. As mesmas rotinas foram utilizadas para depurar e validar o sistema gráfico. / The present work describes the development of a software system to process, visualize and manipulate Magnetic Resonance Images. First an application software W8S developed for an Interpro 32&#176C Intergraph workstation. In this menu and window based 8Oftware, image operations such as filters, transformations and image manipulation based on histograms with statistical data was implemented and evaluated. The knowledge and experience obtained with this implementation was subsequently used to define and design a new graphic environment for a dedicated TMS340 based hardware architecture previously developed in our laboratory. This new toolkit provides window and menu support, and was developed through extensive use of the powerful TMS340 native graphics instructions. Debugging and tracing routines were developed based on an implemented host (PC) - graphics processor board communication system. The same communication software was also used in the examples to perform menu selected downloads of images and functions from the host.

No description available.

Neste trabalho são apresentadas a implementação e avaliação da aplicação do processador gráfico TMS34010. O objetivo foi o desenvolvimento de um sistema para visualização de imagens com este processador, a ser integrado ao Tomógrafo por Ressonância Magnética, do Departamento de Física e Ciência dos Materiais. A fim de avaliar tanto os aspectos do \"hardware\" quanto às capacidades do processador, foram desenvolvidos dois sistemas: um com 16 cores, e outro com 256 níveis de cinza, ambos contendo uma interface serial para interligação de um \"mouse\" convencional para microcomputadores do tipo PC-AT. Através destes sistemas foram avaliados recursos tais como \"pan\" horizontal e vertical, divisão da tela ao meio, vídeo no modo entrelaçado, transparência, operações lógicas, e instruções de transferência de blocos de \"pixels\" / This work describes the implementation and evaluation of a TMS34010 graphics processor based system. The system was developed to be used for the visualization of Magnetic Resonance Images in the MR tomography system under development in our Department. To evaluate the design and the performance of the processor, two systems were developed: one capable to show pictures with 16 simultaneous colors, and a second one able to display pictures with a gray scale of 256 levels. Both systems are equipped with a serial interface supporting a conventional mouse as used for PC compatibles. The hardware developed was used to evaluate implementations of resources such as horizontal and vertical PAN, screen division, interlaced video, transparency, logic operations and pixels block transfer routines

Neste trabalho são apresentadas a implementação e avaliação da aplicação do processador gráfico TMS34010. O objetivo foi o desenvolvimento de um sistema para visualização de imagens com este processador, a ser integrado ao Tomógrafo por Ressonância Magnética, do Departamento de Física e Ciência dos Materiais. A fim de avaliar tanto os aspectos do \"hardware\" quanto às capacidades do processador, foram desenvolvidos dois sistemas: um com 16 cores, e outro com 256 níveis de cinza, ambos contendo uma interface serial para interligação de um \"mouse\" convencional para microcomputadores do tipo PC-AT. Através destes sistemas foram avaliados recursos tais como \"pan\" horizontal e vertical, divisão da tela ao meio, vídeo no modo entrelaçado, transparência, operações lógicas, e instruções de transferência de blocos de \"pixels\" / This work describes the implementation and evaluation of a TMS34010 graphics processor based system. The system was developed to be used for the visualization of Magnetic Resonance Images in the MR tomography system under development in our Department. To evaluate the design and the performance of the processor, two systems were developed: one capable to show pictures with 16 simultaneous colors, and a second one able to display pictures with a gray scale of 256 levels. Both systems are equipped with a serial interface supporting a conventional mouse as used for PC compatibles. The hardware developed was used to evaluate implementations of resources such as horizontal and vertical PAN, screen division, interlaced video, transparency, logic operations and pixels block transfer routines

Purpose of this work is show possibility of using grapichs cart for imaging a video signal. This work is particularly focused on technology CUDA and OpenCL. The solution is first focused on graphics cart and show how has been changed components and how has been changed performaces of graphics cart. Then show CUDA and OpenCL technology itself, and show samples of codes with explain, what which code do. Output of this work is some programs, witch defined for both technology and for both procesors unit. Contribution of this work is show differents between procesors unit, witch can be used to right choose for design your own algorithm.