Examensarbeten för kandidatexamen // Bachelor Theses
Länka till denna samling:
Browse
Browsar Examensarbeten för kandidatexamen // Bachelor Theses efter Ämnesord "Data- och informationsvetenskap"
Visar 1 - 2 av 2
Sökresultat per sida
Sortera efter
- PostEffektiv parallell slumptalsgenerering på GPU:er(2017) Lundberg, Marcus; Aasa, Jakob; Chalmers tekniska högskola / Institutionen för rymd- och geovetenskap; Chalmers University of Technology / Department of Earth and Space SciencesThis work compares several popular pseudo-random number generators implemented on a graphics processing unit (GPU). We consider generation for both uniform and normal distributions. The generators have also been tested using a selection of test algorithms to assess the quality of the output. As a final verification the generators have been tested in-situ on a simulation code. We chose to implement and test five different algorithms for generating uniform distributed numbers and three for generating normal distributed numbers. The generators were implemented with an object oriented programming approach in C++ using Nvidia’s CUDA framework. We have also included generators from Nvidia’s own random number generator library, cuRAND, to compare with our own. The test algorithms were implemented in C++ and CUDA as well. Our results show that some algorithms are not suited for use on GPUs, while other more GPU customized algorithms perform admirably. The results from the simulation code show that all of the generators except Wallace provide good output. The running time of the simulation code is about 100 to 250 times faster on the GPU depending on implementation compared to CPU. From our results we can recommend the Linear Congruental Generator (LCG) for generating uniform numbers if perfomance is a priority, and combining it with the Box-Muller Transform for generating normal distributed numbers.
- PostUppskattning av stjärnbildningstakt hos galaxer i det unga universum(2015) Tengnäs, Viktor; Lan, Richard; Perme, Andreas; Hallqvist, Filippa; Chalmers tekniska högskola / Institutionen för rymd- och geovetenskap; Chalmers University of Technology / Department of Earth and Space SciencesGalaxies in the local universe are almost as old as the universe itself, many of the stars in these galaxies were formed billions of years ago. To get a more complete understanding of the formation and evolution of galaxies, we also need to study younger galaxies in distant regions. This bachelor's thesis uses existing methods in order to study distant galaxies from the time around 3 billion years after the Big Bang, when the star formation rate density is measured to have been at its highest. The aim is to determine the star formation rate and stellar mass of a sample of star-forming galaxies. We study a selection of galaxies of the type sBzK, which are star-forming galaxies at z 1:5 2:5. Interstellar dust is present in these star-forming galaxies, and the dust grains are heated by ultra-violet and optical light from young stars. The dust then re-emits the energy at far-infrared and submillimeter wavelengths. This thermal dust emission can therefore be used to estimate the star formation rate for these galaxies. Data, collected at submillimeter wavelengths, by the interferometric observatory ALMA is analysed and processed to extract the aforementioned properties of the sBzK galaxies. This sample of galaxies are generally too faint to be individually detected at millimeter wavelengths. Therefore two types of stacking methods, image and uv-stacking, are used to estimate their average properties. The results from the different stacking methods are compared with each other. Monte Carlo stacking with both methods is also performed in order to evaluate the results from the stacking processes. With the stacked ALMA data we were able to detect the sBzK sample. The obtained results using uv-stacking was a SFR of 212:3M=yr, when assuming = 1:6, T = 30K and a Chabrier initial mass function. Using image stacking we derive a similar star formation rate of 18 2:4M=yr. Based on simple assumptions we obtain an estimate for the stellar mass of 4;6 1010M. It is also found that the image stacked result gives a lower ux than the uv-stacked result, leading to higher star formation rates for the uv-stacked sources. This star formation rate can be compared to our current knowledge of our own galaxy.