There are many Operation Systems provide the interface likes the frame buffer in Linux. It let Application Programs can read and write the memory block tightly connecting the operating registers of display card directly to get the goal of directly modifying the monitor display. However, although we have the frame buffer, this kind of graphic processing method is not enough to provide a real-time graphic performance under the needed of huge block drawing and moving. In order to eliminate the drawback of the low graphic performance of using pure software, there are many cards with 3D graphic engine produced for speeding up the performance of 3D games, like ATI Radeon X850[1] , NVIDIA GeForce 6800[2] and so on. Although the embedded products like digital TV or mobile phone are not needed to have a complex and powerful 3D graphic engine, the idea of speeding up drafting can be provided for the embedded system as a reference. The graphic engine can not only provide a real-time performance of drafting, but also share the work of CPU in embedded system, to achieve the goal of improving graphic performance and cost down. In the paper, we will implement a 2D graphic engine dynamic shared library for combining the 2D graphic engine and frame buffer on the V/PB926EJ-S target board. To achieve the goal of improving graphic performance, come true the real-time graphic processing ability of Embedded LINUX. And providing a convenient, quickly and reliable software technique of combining hardware resource and operation system together based on the experiment.

A coarse-grain parallelization of SHORECIRC - a quasi-3D nearshore circulation model is implemented. The parallelization is based on the message-passing model for distributed memory architectures using the Message Passing Interface (MPI) standard. The parallel model confirms to the Single Program Multiple Data (SPMD) model. The results from the parallel model have been verified against those from the sequential model for an exact match. The parallel code is portable across different parallel architectures and its performance in terms of speed-up and scalability is studied. A test case simulating rip currents is discussed.

A computacao paralela, empregada no meio cientifico para resolucao de problemas que de- mandam grande poder computacional, teve nos ultimos anos o surgimento de um novo tipo de comunicacao entre instancias do paralelismo. Trata-se da Comunicacao Unilateral (CUL), onde somente uma instancia realiza a operacao de transferencia de informacoes, e esta ocorre em segundo plano, ao contrario da Comunicacao Bilateral (CBL), onde uma instancia envia a informacao e a outra recebe. Neste contexto se buscou analisar os beneficios que a CUL agrega ao paralelismo de um programa que se utiliza de uma grade nao estruturada em me- moria. Duas formas de apoio ao paralelismo foram utilizadas: uma biblioteca, a \"Message Passing Interface\" (MPI) (especificamente a sua parte que descreve a CUL), e uma extensao a linguagem Fortran, o Coarray Fortran (CAF). A semantica do MPI CUL e mais complexa que a do CAF, mas a do CAF e mais restritiva. Para analisar a semantica e desempenho da CUL foi realizada uma ambientacao utilizando MPI CUL e CAF no paralelismo de um programa simples, denominado jogo da Vida (Game of Life), com grade estruturada e com otimo desempenho paralelo atraves do MPI CBL. Na programacao o MPI CUL se mostrou verborragico (aumento do numero de linhas de codigo) e complexo, principalmente quando se utiliza um controle refinado de sincronismo entre as imagens. Ja o CAF reduziu o nu- mero de linhas de codigo (entre 20% e 40%), e o sincronismo e muito mais simples. Os resultados mostraram uma piora no desempenho no caso do MPI CUL, mas para o CAF o desempenho absoluto foi melhor que a implementacao original ate o numero de nucleos de processamento que compartilham a mesma memoria. Para grades nao estruturadas se utilizou o Ocean Land Atmospheric Model (OLAM), um modelo de simulacao do sistema terrestre com grade baseada em prismas triangulares, paralelizado atraves de MPI CBL. A implementacao da comunicacao por MPI CUL na estrutura do paralelismo existente mos- trou que esta semantica possui alguns pontos que podem prejudicar a programacao, como o tratamento da exposicao de memoria (cada instancia tem uma memoria exposta de tamanho diferente) e como e realizado o sincronismo entre as instancias. Em termos de desempenho as curvas de speed-ups mostraram que o MPI CUL prejudicou o OLAM independentemente da implementacao das bibliotecas ou do equipamento utilizado, com reducao de pelo menos 20% no speed-up para sete ou mais processadores. Assim como no jogo da Vida o MPI com comunicacao unilateral penalizou o desempenho. / Parallel computing is used to solve many scientific problems that demand intensive compu- ting power. Recently a new paradigm of communication between instances of the parallelism has appeared, called the one-sided communication (OSC), where only one instance performs the operation of information transfer, occurring in the background, as opposed to the two- sided communication (TSC), where one instance sends the information and the other receives it. In this context we analyze the benefits that OSC aggregates to the parallelism of a pro- gram that uses an unstructured grid in memory. Two OSC implementations were used: the \"Message Passing Interface\" (MPI) library (specifically the part that describes OSC), and Coarray Fortran (CAF), an extension of the Fortran language. The semantics of MPI OSC is more complex than that of CAF, but the semantics of CAF is more restrictive. To analyze the semantics and performance of OSC a simple program called Game of Life is used in a structured grid, giving very good parallel performance through MPI TSC. The MPI OSC program was verbose (increase in the number of lines of code) and complex, especially when using a more refined control to synchronize the parallel instances. On the other hand, CAF has reduced the number of lines of code (between 20% to 40%), and the synchronization is very simple. The results showed a worse performance in the case of MPI OSC, but for the CAF the absolute performance was better than the original implementation up to the number of processor cores that share the same memory. For unstructured grids we used the Ocean Land Atmospheric Model (OLAM), an earth simulation model on a grid based on triangular prisms, and parallelized with MPI TSC. The implementation with MPI OSC showed that this semantics has some points that may affect the coding of the communication structure, as in the treatment of memory exposure (each instance has an exposed memory of different size) and the way to treat the synchronization among instances. In terms of performance, the speedup curves showed that MPI OSC penalized OLAM, independently of the MPI implementation or the equipment used, with a reduction of at least 20% in speedup for seven or more processors. As in the Game of Life, MPI OSC degrades the performance.

A computacao paralela, empregada no meio cientifico para resolucao de problemas que de- mandam grande poder computacional, teve nos ultimos anos o surgimento de um novo tipo de comunicacao entre instancias do paralelismo. Trata-se da Comunicacao Unilateral (CUL), onde somente uma instancia realiza a operacao de transferencia de informacoes, e esta ocorre em segundo plano, ao contrario da Comunicacao Bilateral (CBL), onde uma instancia envia a informacao e a outra recebe. Neste contexto se buscou analisar os beneficios que a CUL agrega ao paralelismo de um programa que se utiliza de uma grade nao estruturada em me- moria. Duas formas de apoio ao paralelismo foram utilizadas: uma biblioteca, a \"Message Passing Interface\" (MPI) (especificamente a sua parte que descreve a CUL), e uma extensao a linguagem Fortran, o Coarray Fortran (CAF). A semantica do MPI CUL e mais complexa que a do CAF, mas a do CAF e mais restritiva. Para analisar a semantica e desempenho da CUL foi realizada uma ambientacao utilizando MPI CUL e CAF no paralelismo de um programa simples, denominado jogo da Vida (Game of Life), com grade estruturada e com otimo desempenho paralelo atraves do MPI CBL. Na programacao o MPI CUL se mostrou verborragico (aumento do numero de linhas de codigo) e complexo, principalmente quando se utiliza um controle refinado de sincronismo entre as imagens. Ja o CAF reduziu o nu- mero de linhas de codigo (entre 20% e 40%), e o sincronismo e muito mais simples. Os resultados mostraram uma piora no desempenho no caso do MPI CUL, mas para o CAF o desempenho absoluto foi melhor que a implementacao original ate o numero de nucleos de processamento que compartilham a mesma memoria. Para grades nao estruturadas se utilizou o Ocean Land Atmospheric Model (OLAM), um modelo de simulacao do sistema terrestre com grade baseada em prismas triangulares, paralelizado atraves de MPI CBL. A implementacao da comunicacao por MPI CUL na estrutura do paralelismo existente mos- trou que esta semantica possui alguns pontos que podem prejudicar a programacao, como o tratamento da exposicao de memoria (cada instancia tem uma memoria exposta de tamanho diferente) e como e realizado o sincronismo entre as instancias. Em termos de desempenho as curvas de speed-ups mostraram que o MPI CUL prejudicou o OLAM independentemente da implementacao das bibliotecas ou do equipamento utilizado, com reducao de pelo menos 20% no speed-up para sete ou mais processadores. Assim como no jogo da Vida o MPI com comunicacao unilateral penalizou o desempenho. / Parallel computing is used to solve many scientific problems that demand intensive compu- ting power. Recently a new paradigm of communication between instances of the parallelism has appeared, called the one-sided communication (OSC), where only one instance performs the operation of information transfer, occurring in the background, as opposed to the two- sided communication (TSC), where one instance sends the information and the other receives it. In this context we analyze the benefits that OSC aggregates to the parallelism of a pro- gram that uses an unstructured grid in memory. Two OSC implementations were used: the \"Message Passing Interface\" (MPI) library (specifically the part that describes OSC), and Coarray Fortran (CAF), an extension of the Fortran language. The semantics of MPI OSC is more complex than that of CAF, but the semantics of CAF is more restrictive. To analyze the semantics and performance of OSC a simple program called Game of Life is used in a structured grid, giving very good parallel performance through MPI TSC. The MPI OSC program was verbose (increase in the number of lines of code) and complex, especially when using a more refined control to synchronize the parallel instances. On the other hand, CAF has reduced the number of lines of code (between 20% to 40%), and the synchronization is very simple. The results showed a worse performance in the case of MPI OSC, but for the CAF the absolute performance was better than the original implementation up to the number of processor cores that share the same memory. For unstructured grids we used the Ocean Land Atmospheric Model (OLAM), an earth simulation model on a grid based on triangular prisms, and parallelized with MPI TSC. The implementation with MPI OSC showed that this semantics has some points that may affect the coding of the communication structure, as in the treatment of memory exposure (each instance has an exposed memory of different size) and the way to treat the synchronization among instances. In terms of performance, the speedup curves showed that MPI OSC penalized OLAM, independently of the MPI implementation or the equipment used, with a reduction of at least 20% in speedup for seven or more processors. As in the Game of Life, MPI OSC degrades the performance.

Wind energy is rapidly growing around the world as non-renewable energy sources are being phased out. To maximize the energy production, it is crucial to construct new wind turbines at locations where the average wind speed is high. An accurate wind assessment of a new potential site is thus essential in the development of new wind farms. Linearized codes (codes based on simplified models) such as Orfeus and WAsP have seen frequent use in the wind industry as a tool for wind resource assessment since they offer reasonable accuracy at a relatively low computational cost. A major issue with the linearized codes on the market is their disregard of stratification, which will cause the codes to omit certain flow characteristics, such as internal gravity waves. Therefore, a new linearized code has been developed, where the modelling of stratification has been included. The developed code has been validated by comparing the simulation results to experimental data found in the literature, where an overall agreement could be found. In addition, test cases for stratified flows over sinusoidal terrain were simulated, showcasing the importance of including stratification as a parameter in the modelling. The developed code in this project may be used for more accurate wind assessment in the future. / Vindenergi växer snabbt världen över då icke-förnybara energikällor håller på att fasas ut. FÖr att maximera energiproduktionen är det viktigt att konstruera vindturbiner där den genomsnittliga vindhastigheten är hög. En noggrann vindbedömning av nya potentiella lägen är därför vitalt i konstruktionen av nya vindparker. Linjäriserade koder (koder baserade på förenklade modeller) som Orefeus and Wasp används ofta i vindindustrin som ett verktyg för vindbedömning, då de erbjuder en hög noggrannhet till en låg beräkningskostad. Ett stort problem med de linjäriserade koderna på marknaden är exkluderingen av stratifikation, vilket får koderna att utelämna vissa flödesegenskaper, som till exempel interna gravitationsvågor. En ny linjäriserad kod har därför utvecklats där modelleringen av stratifikation har inkluderats. Den utvecklade koden har validerats genom att jämföra simuleringarna med experimentella data i litteraturen, där en god överenstämmelse kunda påvisas. Dessutom har testfall av stratifierade flöden över sinusformad terräng genomförts, vilket påvisade betydelsen av att inkludera stratifikation som en parameter i modelleringen. Den utvecklade koden i detta projekt kan användas för en mer representativ vindbedömning i framtiden.

O presente trabalho descreve a implementação do algoritmo discreto da FFT-2D, numa rede de transputers. Primeiramente a implementação seqüencial do algoritmo é analisado, em seguida são apresentados algumas técnicas de paralelização, bem como sua aplicação no algoritmo da FFT-2D. Finalmente são apresentados os resultados do desempenho obtido por redes compostas de um, dois e quatro transputers / The present work describes the implementation of the discrete FFT-2D algorithm in a distributed transputer network. First a seqüencial implementation of the algorithm is presented. Then some parallelization techniques are analyzed and applied to the FFT-2D algorithm. Finally the obtained performance is presented for networks containing one, two and four transputers

A formação do perfil de velocidades do vento é influenciada pela rugosidade e topografia do terreno; em um terreno plano, por exemplo, as velocidades são diminuídas conforme a sua proximidade com a superfície terrestre, determinando a constituição da camada limite atmosférica (C.L.A.). O escoamento do vento em terrenos complexos, ou seja, terrenos que possuem morros e taludes, sejam isolados ou múltiplos, possui o perfil de velocidades modificado, fazendo com que para cotas mais próximas da superfície se observe um aumento das velocidades. A este incremento de velocidades dá-se o nome de speed-up. Esta pesquisa tem como foco a investigação da estrutura do escoamento do vento, analisando minuciosamente o perfil de velocidades em um escoamento turbulento, para topografias isoladas e complexas. Para tanto, foram utilizados dois métodos para obtenção do perfil de velocidades: ensaios experimentais em túnel de vento e aplicação de normas e modelos analítcos de carregamento do vento. O plano experimental engloba nove topografias dentre as quais estão: quatro morros bidimensionais (2D) isolados, quatro morros tridimensionais (3D) isolados e um morro 3D de uma topografia complexa. Os modelos foram ensaiados para duas categorias de terreno, conforme a Norma Brasileira NBR 6123 (ABNT, 1988): I – superfícies lisas de grandes dimensões e III-IV – terreno coberto por obstáculos, no túnel de vento Prof° Joaquim Blessmann. Após a execução do plano experimental aplicou-se modelos analíticos com as mesmas parametrizações do modelo experimental e por fim foram comparados entre si. Os modelos analíticos estudados foram: Jackson e Hunt (1975, Lemelin, Surry e Davenport (1988) e as normas estudadas são: NBR 6123 (ABNT, 1988), Eurocode 1 (CEN-TC, 2010), AIJ (AIJ, 2004), NBCC (NRCC, 2010), AS/NZS (AS/NZS, 2011), ASCE (ASCE, 2010). O estudo comparativo entre os modelos analíticos e normas revela a inexistência de uma homogeneidade entre os modelos, visto que a formulação de cálculo é distinta, quando temos uma topografia em análise. Além disso, os modelos analíticos, quando comparados com os ensaios experimentais, tendem a ser mais conservadores tanto para modelos 2D ou 3D, para pontos à barlavento, no cume e à sotavento do morro. A NBR 6123 apresentou os maiores incrementos de velocidades em comparação com os demais modelos e com os dados experimentais. / The formation of the wind speed profile is influenced by the roughness and topography of the ground; on a flat ground, for example, the speeds are reduced according to their proximity to the earth’s surface, determining the constitution of the atmospheric boundary layer (C.L.A.). The wind flow in complex terrain, in other words, whether single or multiples, has the modified wind speed profile, causing closest to the surface dimensions is observed increased speeds. To this increase speeds gives the name of speed-up. This research focuses on the investigation of the wind flow structure, thoroughly analyzing the speeds profile in turbulent flow, for isolated and complex topographies. For this purpose, two methods for obtaining the speed profile were used: experimental tests in a wind tunnel and application standards or wind loading codes. The experimental plan includes nine topographies among which are: four 2D hills (2D) isolates, four 3D hills (3D) isolates and a complex topography hill. The models were tested for two roughness according to the Brazilian standard NBR 6123 (ABNT, 1988): I – smooth surfaces large dimensions and III-IV – ground covered by obstacles, in the wind tunnel Prof° Joaquim Blessmann. After the execution of the experimental tests were applied analytical models with the same parametrization of the experimental model and finally were compared. The analytical models studied were: Jackson e Hunt (1975, Lemelin, Surry e Davenport (1988), NBR 6123 (ABNT, 1988), Eurocode 1 (CEN-TC, 2010), AIJ (AIJ, 2004), NBCC (NRCC, 2010), AS/NZS (AS/NZS, 2011), ASCE (ASCE, 2010. The compararative study between the analytical models and standards reveals the lack of homogeneity between the models, since the calculation formulation is different when we have a topography analysis. Furthermore, analytical models, compared with the experimental tests tend to be more conservative for both 2D and 3D models, points to the windward, top of the hill and leeward. The NBR 6123 showed the largest increases in speed compared to the other models and experimental data.

In today’s machine learning research, probabilistic graphical models are used extensively to model complicated systems with uncertainty, to help understanding of the problems, and to help inference and predict unknown events. For inference tasks, exact inference methods such as junction tree algorithms exist, but they suffer from exponential growth of cluster size and thus is not able to handle large and highly connected graphs. Approximate inference methods do not try to find exact probabilities, but rather give results that improve as algorithm runs. Gibbs sampling, as one of the approximate inference methods, has gained lots of traction and is used extensively in inference tasks, due to its ease of understanding and implementation. However, as problem size grows, even the faster algorithm needs a speed boost to meet application requirement. The number of variables in an application graphical model can range from tens of thousands to billions, depending on problem domain. The original sequential Gibbs sampling may not return satisfactory result in limited time. Thus, in this thesis, we investigate in ways to speed up Gibbs sampling. We will study ways to do better initialization, blocking variables to be sampled together, as well as using simulated annealing. These are the methods that modifies the algorithm itself. We will also investigate in ways to parallelize the algorithm. An algorithm is parallelizable if some steps do not depend on other steps, and we will find out such dependency in Gibbs sampling. We will discuss how the choice of different hardware and software architecture will affect the parallelization result. We will use optical flow problem as an example to demonstrate the various speed up methods we investigated. An optical flow method tries to find out the movements of small image patches between two images in a temporal sequence. We demonstrate how we can model it using probabilistic graphical model, and solve it using Gibbs sampling. The result of using sequential Gibbs sampling is demonstrated, with comparisons from using various speed up methods and other optical flow methods.

A formação do perfil de velocidades do vento é influenciada pela rugosidade e topografia do terreno; em um terreno plano, por exemplo, as velocidades são diminuídas conforme a sua proximidade com a superfície terrestre, determinando a constituição da camada limite atmosférica (C.L.A.). O escoamento do vento em terrenos complexos, ou seja, terrenos que possuem morros e taludes, sejam isolados ou múltiplos, possui o perfil de velocidades modificado, fazendo com que para cotas mais próximas da superfície se observe um aumento das velocidades. A este incremento de velocidades dá-se o nome de speed-up. Esta pesquisa tem como foco a investigação da estrutura do escoamento do vento, analisando minuciosamente o perfil de velocidades em um escoamento turbulento, para topografias isoladas e complexas. Para tanto, foram utilizados dois métodos para obtenção do perfil de velocidades: ensaios experimentais em túnel de vento e aplicação de normas e modelos analítcos de carregamento do vento. O plano experimental engloba nove topografias dentre as quais estão: quatro morros bidimensionais (2D) isolados, quatro morros tridimensionais (3D) isolados e um morro 3D de uma topografia complexa. Os modelos foram ensaiados para duas categorias de terreno, conforme a Norma Brasileira NBR 6123 (ABNT, 1988): I – superfícies lisas de grandes dimensões e III-IV – terreno coberto por obstáculos, no túnel de vento Prof° Joaquim Blessmann. Após a execução do plano experimental aplicou-se modelos analíticos com as mesmas parametrizações do modelo experimental e por fim foram comparados entre si. Os modelos analíticos estudados foram: Jackson e Hunt (1975, Lemelin, Surry e Davenport (1988) e as normas estudadas são: NBR 6123 (ABNT, 1988), Eurocode 1 (CEN-TC, 2010), AIJ (AIJ, 2004), NBCC (NRCC, 2010), AS/NZS (AS/NZS, 2011), ASCE (ASCE, 2010). O estudo comparativo entre os modelos analíticos e normas revela a inexistência de uma homogeneidade entre os modelos, visto que a formulação de cálculo é distinta, quando temos uma topografia em análise. Além disso, os modelos analíticos, quando comparados com os ensaios experimentais, tendem a ser mais conservadores tanto para modelos 2D ou 3D, para pontos à barlavento, no cume e à sotavento do morro. A NBR 6123 apresentou os maiores incrementos de velocidades em comparação com os demais modelos e com os dados experimentais. / The formation of the wind speed profile is influenced by the roughness and topography of the ground; on a flat ground, for example, the speeds are reduced according to their proximity to the earth’s surface, determining the constitution of the atmospheric boundary layer (C.L.A.). The wind flow in complex terrain, in other words, whether single or multiples, has the modified wind speed profile, causing closest to the surface dimensions is observed increased speeds. To this increase speeds gives the name of speed-up. This research focuses on the investigation of the wind flow structure, thoroughly analyzing the speeds profile in turbulent flow, for isolated and complex topographies. For this purpose, two methods for obtaining the speed profile were used: experimental tests in a wind tunnel and application standards or wind loading codes. The experimental plan includes nine topographies among which are: four 2D hills (2D) isolates, four 3D hills (3D) isolates and a complex topography hill. The models were tested for two roughness according to the Brazilian standard NBR 6123 (ABNT, 1988): I – smooth surfaces large dimensions and III-IV – ground covered by obstacles, in the wind tunnel Prof° Joaquim Blessmann. After the execution of the experimental tests were applied analytical models with the same parametrization of the experimental model and finally were compared. The analytical models studied were: Jackson e Hunt (1975, Lemelin, Surry e Davenport (1988), NBR 6123 (ABNT, 1988), Eurocode 1 (CEN-TC, 2010), AIJ (AIJ, 2004), NBCC (NRCC, 2010), AS/NZS (AS/NZS, 2011), ASCE (ASCE, 2010. The compararative study between the analytical models and standards reveals the lack of homogeneity between the models, since the calculation formulation is different when we have a topography analysis. Furthermore, analytical models, compared with the experimental tests tend to be more conservative for both 2D and 3D models, points to the windward, top of the hill and leeward. The NBR 6123 showed the largest increases in speed compared to the other models and experimental data.

A formação do perfil de velocidades do vento é influenciada pela rugosidade e topografia do terreno; em um terreno plano, por exemplo, as velocidades são diminuídas conforme a sua proximidade com a superfície terrestre, determinando a constituição da camada limite atmosférica (C.L.A.). O escoamento do vento em terrenos complexos, ou seja, terrenos que possuem morros e taludes, sejam isolados ou múltiplos, possui o perfil de velocidades modificado, fazendo com que para cotas mais próximas da superfície se observe um aumento das velocidades. A este incremento de velocidades dá-se o nome de speed-up. Esta pesquisa tem como foco a investigação da estrutura do escoamento do vento, analisando minuciosamente o perfil de velocidades em um escoamento turbulento, para topografias isoladas e complexas. Para tanto, foram utilizados dois métodos para obtenção do perfil de velocidades: ensaios experimentais em túnel de vento e aplicação de normas e modelos analítcos de carregamento do vento. O plano experimental engloba nove topografias dentre as quais estão: quatro morros bidimensionais (2D) isolados, quatro morros tridimensionais (3D) isolados e um morro 3D de uma topografia complexa. Os modelos foram ensaiados para duas categorias de terreno, conforme a Norma Brasileira NBR 6123 (ABNT, 1988): I – superfícies lisas de grandes dimensões e III-IV – terreno coberto por obstáculos, no túnel de vento Prof° Joaquim Blessmann. Após a execução do plano experimental aplicou-se modelos analíticos com as mesmas parametrizações do modelo experimental e por fim foram comparados entre si. Os modelos analíticos estudados foram: Jackson e Hunt (1975, Lemelin, Surry e Davenport (1988) e as normas estudadas são: NBR 6123 (ABNT, 1988), Eurocode 1 (CEN-TC, 2010), AIJ (AIJ, 2004), NBCC (NRCC, 2010), AS/NZS (AS/NZS, 2011), ASCE (ASCE, 2010). O estudo comparativo entre os modelos analíticos e normas revela a inexistência de uma homogeneidade entre os modelos, visto que a formulação de cálculo é distinta, quando temos uma topografia em análise. Além disso, os modelos analíticos, quando comparados com os ensaios experimentais, tendem a ser mais conservadores tanto para modelos 2D ou 3D, para pontos à barlavento, no cume e à sotavento do morro. A NBR 6123 apresentou os maiores incrementos de velocidades em comparação com os demais modelos e com os dados experimentais. / The formation of the wind speed profile is influenced by the roughness and topography of the ground; on a flat ground, for example, the speeds are reduced according to their proximity to the earth’s surface, determining the constitution of the atmospheric boundary layer (C.L.A.). The wind flow in complex terrain, in other words, whether single or multiples, has the modified wind speed profile, causing closest to the surface dimensions is observed increased speeds. To this increase speeds gives the name of speed-up. This research focuses on the investigation of the wind flow structure, thoroughly analyzing the speeds profile in turbulent flow, for isolated and complex topographies. For this purpose, two methods for obtaining the speed profile were used: experimental tests in a wind tunnel and application standards or wind loading codes. The experimental plan includes nine topographies among which are: four 2D hills (2D) isolates, four 3D hills (3D) isolates and a complex topography hill. The models were tested for two roughness according to the Brazilian standard NBR 6123 (ABNT, 1988): I – smooth surfaces large dimensions and III-IV – ground covered by obstacles, in the wind tunnel Prof° Joaquim Blessmann. After the execution of the experimental tests were applied analytical models with the same parametrization of the experimental model and finally were compared. The analytical models studied were: Jackson e Hunt (1975, Lemelin, Surry e Davenport (1988), NBR 6123 (ABNT, 1988), Eurocode 1 (CEN-TC, 2010), AIJ (AIJ, 2004), NBCC (NRCC, 2010), AS/NZS (AS/NZS, 2011), ASCE (ASCE, 2010. The compararative study between the analytical models and standards reveals the lack of homogeneity between the models, since the calculation formulation is different when we have a topography analysis. Furthermore, analytical models, compared with the experimental tests tend to be more conservative for both 2D and 3D models, points to the windward, top of the hill and leeward. The NBR 6123 showed the largest increases in speed compared to the other models and experimental data.