Valutazione delle prestazioni di processori a basso consumo energetico in applicazioni parallele

In questo lavoro di tesi sono state impiegate le librerie grafiche OpenGL ES 2 per eseguire calcoli paralleli sulla GPU del Raspberry Pi. Sono stati affrontati e discussi concetti riguanrdati il calcolo parallelo, stream processing, GPGPU e le metriche di valutazione di algoritmi paralleli. Sono inoltre descritte le potenzialita e le limitazioni derivanti dall'impiego di OpenGL per implementare algoritmi paralleli. In particolare si e fatto riferimento all'algoritmo Seam Carving per il restringimento di immagini, realizzando e valutando una implementazione parallela di questo sul Raspberry Pi.

Una libreria OpenGL per la selezione e editing di mesh poligonali

Nel mondo Open Source, la libreria grafica OpenGL è oggi ampiamente utilizzata in svariati settori come l'animazione 2D/3D, la modellazione CAD o nello sviluppo di videogiochi. A causa dei suoi innumerevoli usi e dell'astrazione che OpenGL permette di ottenere su diversi ambienti grafici, lo sviluppatore - che la utilizza - è vincolato a cercare librerie di supporto al fine di sfruttarne al meglio le potenzialità. Questa tesi si configura su questi presupposti, presentando una libreria di selezione e editing di mesh 3D basata su OpenGL. La libreria, chiamata libEditMesh, sfrutta il meccanismo geometrico del RayPicking permettendo all'utilizzatore di identificare col mouse punti, facce e lati di solidi in scena. La tesi si articola sostanzialmente in due parti: nella prima vengono proposte alcune soluzioni ad-hoc sviluppate su applicazioni già esistenti nel panorama openSource, e non; nella seconda vengono esposti gli algoritmi e funzioni implementate in libEditMesh.

Visualization of complex biological systems: An immune response model using OpenGL

In this paper we present an update on our novel visualization technologies based on cellular immune interaction from both large-scale spatial and temporal perspectives. We do so with a primary motive: to present a visually and behaviourally realistic environment to the community of experimental biologists and physicians such that their knowledge and expertise may be more readily integrated into the model creation and calibration process. Visualization aids understanding as we rely on visual perception to make crucial decisions. For example, with our initial model, we can visualize the dynamics of an idealized lymphatic compartment, with antigen presenting cells (APC) and cytotoxic T lymphocyte (CTL) cells. The visualization technology presented here offers the researcher the ability to start, pause, zoom-in, zoom-out and navigate in 3-dimensions through an idealised lymphatic compartment.

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.

Arduino plataforma mugikor baten erabilera RaspberryPi eta ROS bitartez

Txosten honetan azaltzen den proiektua, auto itxurako robot baten eraikuntzan oinarritzen da. Arduino plataforman oinarritutako robot mugikor baten sorkuntza burutuko da hutsetik, honen kontrola Raspberry Pi ordenagailu txikiaren bitartez eginez. Gainera, azken gailu honi ahalmen handiagoa emateko asmoz, ROS plataforma instalatuko da bertan. Duten kostu baxua dela eta, gaur egun izugarrizko arrakasta lortu dute plataforma guzti hauek, baina proiektu honetan beraien ahalmena neurtu nahi dugu ezaugarri zehatz batzuk dituen robota sortuz.

基于OpenGL的真空机器人三维图形仿真

真空机器人是先进半导体制造行业中不可缺少的重要设备之一。本文介绍了一种并联型真空机器人的运动学模型,以及利用OpenGL和STL格式文件在Visual C++6.0程序界面中实现3D图形仿真的方法,取得了良好的实验效果,为进一步实现上位机实时在线监控奠定了基础。

可重构星球探测机器人的OpenGL仿真实验平台研究

采用基于Visual C++和OpenGL的建模和运动仿真方法,对可重构星球探测机器人系统的三维运动仿真实验平台进行了研究,建立了一个多机器人系统的仿真实验平台。开发的实验平台可用于探索和验证机器人系统的工作原理、工作空间、多机器人协调算法、重构方法、系统集成技术等。在该平台上进行了机器人的运动学仿真和协调运动研究,验证了该仿真平台的有效性和机器人系统体系结构的合理性。

基于OpenGL的交会对接仿真系统

针对基于视觉的空间载体位姿测量很难进行真实实验的特点,提出了一种基于OpenG1的半物理仿真实验方法。该方法通过计算机图形学技术和计算机视觉技术的结合,能够直观、快速地模拟空间载体位姿测量过程。

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.

Estudio de los estándares gráficos actuales : desarrollo de un videojuego de plataformas utilizando las tecnologías OpenGL y DirectX

[ES]Implementación de un video juego de lucha y plataformas para 2 a 4 jugadores, basada en la serie de juegos "Super Smash Bros" de Nintendo. Se selecciona el prototipo más viable y se termina de programar el juego con este prototipo. Cada jugador controla a un personaje. Los personajes luchan sobre un conjunto de plataformas para lanzar a sus oponentes. El juego termina cuando solo queda un personaje con vida. Los personajes se mueven en tres dimensiones. El proyecto se desarrolla en Visual C++ Express, con programación orientada a objetos. Utiliza DirectInput y XInput para las entradas de datos.

Revisione critica dei risultati e nuovi algoritmi decisionali sulla chirurgia dell'OSAS

Obstructive sleep apnoea/hypopnoea syndrome (OSAHS) is the periodic reduction or cessation of airflow during sleep. The syndrome is associated whit loud snoring, disrupted sleep and observed apnoeas. Surgery aims to alleviate symptoms of daytime sleepiness, improve quality of life and reduce the signs of sleep apnoea recordered by polysomnography. Surgical intervention for snoring and OSAHS includes several procedures, each designed to increase the patency of the upper airway. Procedures addressing nasal obstruction include septoplasty, turbinectomy, and radiofrequency ablation (RF) of the turbinates. Surgical procedures to reduce soft palate redundancy include uvulopalatopharyngoplasty with or without tonsillectomy, uvulopalatal flap, laser-assisted uvulopalatoplasty, and RF of the soft palate. More significant, however, particularly in cases of severe OSA, is hypopharyngeal or retrolingual obstruction related to an enlarged tongue, or more commonly due to maxillomandibular deficiency. Surgeries in these cases are aimed at reducing the bulk of the tongue base or providing more space for the tongue in the oropharynx so as to limit posterior collapse during sleep. These procedures include tongue-base suspension, genioglossal advancement, hyoid suspension, lingualplasty, and maxillomandibular advancement. We reviewed 269 patients undergoing to osas surgery at the ENT Department of Forlì Hospital in the last decade. Surgery was considered a success if the postoperative apnea/hypopnea index (AHI) was less than 20/h. According to the results, we have developed surgical decisional algorithms with the aims to optimize the success of these procedures by identifying proper candidates for surgery and the most appropriate surgical techniques. Although not without risks and not as predictable as positive airway pressure therapy, surgery remains an important treatment option for patients with obstructive sleep apnea (OSA), particularly for those who have failed or cannot tolerate positive airway pressure therapy. Successful surgery depends on proper patient selection, proper procedure selection, and experience of the surgeon. The intended purpose of medical algorithms is to improve and standardize decisions made in the delivery of medical care, assist in standardizing selection and application of treatment regimens, to reduce potential introduction of errors. Nasal Continuous Positive Airway Pressure (nCPAP) is the recommended therapy for patients with moderate to severe OSAS. Unfortunately this treatment is not accepted by some patient, appears to be poorly tolerated in a not neglible number of subjects, and the compliance may be critical, especially in the long term if correctly evaluated with interview as well with CPAP smart cards analysis. Among the alternative options in Literature, surgery is a long time honoured solution. However until now no clear scientific evidence exists that surgery can be considered a really effective option in OSAHS management. We have design a randomized prospective study comparing MMA and a ventilatory device (Autotitrating Positive Airways Pressure – APAP) in order to understand the real effectiveness of surgery in the management of moderate to severe OSAS. Fifty consecutive previously full informed patients suffering from severe OSAHS were enrolled and randomised into a conservative (APAP) or surgical (MMA) arm. Demographic, biometric, PSG and ESS profiles of the two group were statistically not significantly different. One year after surgery or continuous APAP treatment both groups showed a remarkable improvement of mean AHI and ESS; the degree of improvement was not statistically different. Provided the relatively small sample of studied subjects and the relatively short time of follow up, MMA proved to be in our adult and severe OSAHS patients group a valuable alternative therapeutical tool with a success rate not inferior to APAP.