This thesis studies the impact of hardware features of graphics cards on performance of GPU computing using GPGPU-Sim simulation software tool. GPU computing is a growing topic in the world of computing, and could be an important milestone for computers. Therefore, such a study that seeks to identify the performance bottlenecks of the program with respect to hardware parameters of the devvice can be considered an important step towards tuning devices for higher efficiency. In this work we selected convolution algorithm - a typical GPGPU application - and conducted several tests to study different performance parameters. These tests were performed on two simulated graphics cards (NVIDIA GTX480, NVIDIA Tesla C2050), which are supported by GPGPU-Sim. By changing the hardware parameters of graphics card such as memory cache sizes, frequency and the number of cores, we can make a fine-grained analysis on the effect of these parameters on the performance of the program. A graphics card working on a picture convolution task releis on the L1 cache but has the worst performance with a small shared memory. Using this simulator to run performance tests on a theoretical GPU architecture could lead to better GPU design for embedded systems.

Stuttgart, Univ., Studienarb., 2002.

Anterior cruciate ligaments (ACL) are among the most common reported ligament injuries in athletes. This injury has been linked to changes in joint stability, neuromuscular activity and contact mechanics. In vitro simulators have proven to provide valuable insights on the potential effects of muscle activity on joint stability. The University of Ottawa Knee Simulate (UOKS) is a mechanical load driven quasi-static apparatus that provides the framework to explore pressure changes in knee compartments and the resultant kinematics in response to six individual muscle loads applied around an unconstrained joint. The main objective of the present work was to develop an automated loading mechanism for the UOKS. Furthermore, the secondary objective was to evaluate the accuracy, precision, reliability and validity of the newly automated system. Six transmission units were designed, built and tested as part of the development of the automated mechanism adapted to the UOKS. Load cells were used to obtain real time feedback of the load created by the transmission units. A software controller was programmed using LABVIEW to control these transmissions as a graphic user interface (GUI). Each transmission was tested independently and compared to an external master load cell. Furthermore, four cadaveric knee joints were mounted and suspended inside the UOKS for experimentation. Seven different loading conditions were tested with the ACL intact and after the ACL was severed. Pressure and kinematic data were recorded to correlate the changes in these variables due to changes in loading conditions simulated by the UOKS. The controlled loading experiment of the automated mechanism showed the accuracy of the controller to be within +/-1N, and multiple trials showed the system’s capability to produce loads. Additionally, the results showed the controller to have an ICC of 0.99 between the load produced and the target load. The comparison between feedback load cells and the master load cell displayed unique results regarding each transmission and the loads being applied. The results showed overestimations and underestimations with unique load difference trends for each transmission. Nevertheless, the relative difference measured by the master load cell was below 10% in all the transmissions. This study demonstrated that the controller was accurate and reliable when producing loads. On the other hand, the results of the integration analysis showed the presence of losses in the system when transmitting the loads from the actuators to the front of the UOKS. These losses were different for each transmission, likely due to the independence of each pulley system. Nevertheless, the automated loading mechanism proved to be a valid replacement for the original static loading mechanism previously used by the UOKS and is capable of reliably simulating six independent loading conditions.

Introduction: Understanding the relationship between muscle loads crossing the knee joint and knee joint mechanics is critical for understanding knee stability and the effects of altered muscle forces on healthy and ACL injured knees. In vitro measurement can be used to elucidate this if the simulation is biofidelic, allowing the physiological levels of applied loads to dictate the tibiofemoral kinematics in all degrees of freedom (DoF). The objectives of this study were to describe and apply the University of Ottawa knee simulator as well as measure the reliability of the device. In addition, this device was used to quantify the effect of muscle loads and anterior cruciate ligament (ACL) resection on contact mechanics and kinematics of the tibiofemoral joint. Methods: Muscle forces were determined from an electromyography-driven musculoskeletal model of a healthy male during gait. Six knee specimens were loaded into the simulator and subjected to 100%, 75% and 50% in vivo muscle forces applied through the 6 simulated muscles, in addition to a quadriceps weakness and a hamstring weakness condition. Tibiofemoral mechanics were measured with all 5 loading conditions before and after ACL transection. Results: With the ACL intact, very high reliability in contact area and pressures among loading conditions were observed as the intra-class correlation coefficients (ICC) ranged from 0.932 to 0.99. After ACL transection, reliability remained very high as ICCs ranged from 0.926 to 0.99. In all simulated conditions, muscle forces maintained the knee joint in a stable position resulting in minimal kinematic differences, but altered contact mechanics in both the ACL and non-ACL condition. Removal of the ACL significantly reduced both the medial and lateral contact areas in all loading conditions compared to the ACL intact condition. Conclusion: In summary, the UOKS has demonstrated high reliability within repeated measures. Additionally, small, normally undetectable alterations in joint kinematics resulted in significant alterations to contact mechanics, which can be linked to the degenerative process.

This master thesis is focused on description of the standards for mobile and portable receive of a digital television broadcast DVBT2Lite and DVBNGH. In MATLAB has been developed an application simulator with adjustable parameters of the transmitter and transmission channel. Application shows the constellation diagram of the received signal, resultant bit error and frequency characteristic of transmission channel. In this thesis are discussed simulation results with measured results from different simulation scenarios.

With the use of tow-tank experiments, data may be generated for ships of various classes using comprehensive instrumentation. This data gives the ability to determine the response of ships to various sea-state conditions far in advanced of their construction and launch. However, this data does not indicate the effects of those sea-states to the individuals aboard that ship. In order to define these effects a sea-state simulator must be designed and built. Once construction is completed a series of test must be conducted to determine the response of the simulator. This response allows the comparison to actual tow-tank data to determine if the simulator is capable of performing the desired research. / Master of Science

No description available.

The growing performance gap between CPUs and sustainable disk I/O is a major hurdle in supporting modern applications. As the CPUs become faster, this gap is projected to worsen, thus making it a critical problem that should be addressed with high priority. Although efficient algorithms have alleviated this problem, the mechanical nature of the disk places physical limits on the achievable speedup. On the other hand, newer technologies such as flash memory promise significant improvements in access time, power consumption, and storage density. However, the mature disk technology offers the most favorable cost per bit ratio. Since replacing standard hard disks with flash disks is prohibitively expensive, hybrid hard disks augment the standard hard disk with a small amount of flash memory. By exploiting the beneficial aspects of both technologies they aim to provide breakthrough increase in performance. Nevertheless, hybrid hard disks pose several significant design challenges. Effective and efficient algorithms to manage the flash, the disk, and interaction between them are required. To facilitate rapid and easy exploration of the design space for hybrid hard disk algorithms we present the design and implementation of a flexible and extensible simulator that models hybrid hard disks. The simulator is flexible in that it models several configurations in which the flash and the magnetic medium interact. The simulator is extensible in that it provides a simple framework to plug in several algorithms to manage the hybrid hard disk. We validate our simulator and analyze the performance of the hybrid hard disk for real workloads. / Master of Science

In this world of fast-paced learning, training agencies often require their learners to acquire the knowledge and skills needed for a job at an expedited rate. Because of this rapid form of training, learners are sometimes uncertain about their abilities to execute task-based performances. This uncertainty can lead to a decrease in learners’ self-efficacy on expected task performance. In order to help with this training, trainers are using a variety of simulations and simulators to provide learners’ valuable and necessary training experiences. This mixed methods study explored the influence of high-fidelity virtual training simulators on learners’ self-efficacy. It used pre- and post-simulation-use surveys that combined general self-efficacy questions (Schwarzer & Jerusalem, 1995) and task-specific self-efficacy questions (Bandura, 1977, 1997, 2006; Bandura, Adams, Hardy, & Howells, 1980). This study had a sample size of 18 participants. It was assumed that the intent of providing learners with the vital experience needed to perform specific tasks in a high-fidelity virtual training simulator was to increase their self-efficacy on task-specific criteria. Instead, through surveys, observations, and interviews, the research revealed a decrease in learners’ self-efficacy due to heightened emotional arousal stemming from the learners’ experiences with the level of realism the simulator provide, as well as with breakdowns within the simulator. The breakdowns and the realism were the most influential aspects that influenced self-efficacy in this study. The significance of these findings shows that despite learners wanting to use high-fidelity virtual training simulators, improperly functioning simulators can negatively influence learners’ self-efficacy in task-based performances. / Ph. D.

In this thesis, an Interactive Video Data Service (IVDS) is developed. This service provides a mechanism for television viewers to interact with the program they are watching. Possible interactions include purchasing products from home shopping programs and requesting information from advertisers. Within the project, two areas were focused upon: channel simulation and the Repeater Unit. Additionally, the overall system was discussed along with its background. The purpose of channel simulation was to demonstrate the viability of the unique communication channel model proposed for the IVDS system. This channel was implemented using uni-directional transmissions, without acknowledgments. The Repeater Unit was designed to be a message processing system, intended to relay messages from system users to the home office. The design entailed both hardware and software. The hardware requirements were for a high level design, while the software required not only design, but implementation. / Master of Science