We conduct a comprehensive and systematic study of the Fermi-Pasta-Ulam (FPU) model using both equilibrium and non-equilibrium molecular dynamics simulations, with the aim being to explain the cause of the anomalous energy-transport behaviour in the model. In the equilibrium scenario, our motivation stems from the lack of a complete understanding of the effects of initial conditions on the energy dissipation among Fourier modes. We also critically reconsider the ????probes' that had been widely used to quantitatively describe the types of energy sharing in a system, and then decide on a preferred choice to be used in our equilibrium study. We establish, from strong numerical evidence, that there exists a critical energy density of approximately 0:1, above which the energy dissipation among the modes becomes independent of initial conditions and system parameters, and that the full equipartition of mode energy is never attained in the FPU model. We report, for the first time, the violation of particle positions in the FPU model at high energies, where the particles are found to pass through one another. In the non-equilibrium scenario, we critically review the Nos???Se-Hoover algorithm thermostatting method largely used by other works, and identify its weaknesses. We also review some other alternative methods and decide on the most appropriate one to be implemented throughout our work. We confirm the divergence of the thermal conductivity of the FPU model as the chain length increases, and that kfpu [symbol] No.41, in agreement with other works. Our study further shows that there exists an upper limit of the anharmonicity in the FPU model, and that any attempt to increase the strength of this anharmonicity will not succeed. We also introduce elastic collisions into the original FPU model and find that the Modified model (FPUC) still exhibits anomalous thermal conductivity. We conclude that a one-dimensional FPU-type model with ????only' nearest-neighbour interaction, regardless of being soft or hard, does not exhibit a finite thermal conductivity as the system size increases, due to the non-chaotic nature of its microscopic dynamics, the origin of which we are unable to account for. Finally, we briefly outline possible research directions.

We conduct a comprehensive and systematic study of the Fermi-Pasta-Ulam (FPU) model using both equilibrium and non-equilibrium molecular dynamics simulations, with the aim being to explain the cause of the anomalous energy-transport behaviour in the model. In the equilibrium scenario, our motivation stems from the lack of a complete understanding of the effects of initial conditions on the energy dissipation among Fourier modes. We also critically reconsider the ????probes' that had been widely used to quantitatively describe the types of energy sharing in a system, and then decide on a preferred choice to be used in our equilibrium study. We establish, from strong numerical evidence, that there exists a critical energy density of approximately 0:1, above which the energy dissipation among the modes becomes independent of initial conditions and system parameters, and that the full equipartition of mode energy is never attained in the FPU model. We report, for the first time, the violation of particle positions in the FPU model at high energies, where the particles are found to pass through one another. In the non-equilibrium scenario, we critically review the Nos???Se-Hoover algorithm thermostatting method largely used by other works, and identify its weaknesses. We also review some other alternative methods and decide on the most appropriate one to be implemented throughout our work. We confirm the divergence of the thermal conductivity of the FPU model as the chain length increases, and that kfpu [symbol] No.41, in agreement with other works. Our study further shows that there exists an upper limit of the anharmonicity in the FPU model, and that any attempt to increase the strength of this anharmonicity will not succeed. We also introduce elastic collisions into the original FPU model and find that the Modified model (FPUC) still exhibits anomalous thermal conductivity. We conclude that a one-dimensional FPU-type model with ????only' nearest-neighbour interaction, regardless of being soft or hard, does not exhibit a finite thermal conductivity as the system size increases, due to the non-chaotic nature of its microscopic dynamics, the origin of which we are unable to account for. Finally, we briefly outline possible research directions.

The technology in cellular phones, portable computing systems, intelligent- andconnected- devices are evolving in a high pace and in many cases these devices arerequired to operate in a low-power environment. The problem that continues toemerge, is the power consumption in microcontrollers and DSP devices. This issue hasover time become important to solve in order to maximize battery life. To ease thechoice of power efficient microcontrollers, controlled experiments were thereforeperformed with the Cortex-M4, this microcontroller was chosen because of theupgraded hardware, which has led to an appreciable change in both power- and speedefficiency compared to its predecessors.The conclusion presents important points, along with advantages and difficulties toconsider when implementing a DSP application. By comparing different optimizationswith the Floating Point Unit(FPU), Fixed-point and software Floating-point, the resultsshow that there are major differences in power consumption between these threeoptions. Depending on which option and optimization used then the powerconsumption can exceed over 70% more compared to the other options available.

No description available.

This thesis deals with the design of a floating point module, which allows to perform operations with floating point operands that have any bit width. For this purpose, the module is implemented as a template class in C ++. The module is designed to allow it to be used when designing an application-specific processor. First, the floating point number and template functions in c ++ are described. In the practical part the algorithms of the individual operations and the design of the module itself are described as template libraries.