high performance computing on graphics processing units: hgpu.org

The goal of this dissertation was to parallelize a dose calculation code for radiotherapy cancer treatment and explore the suitability of the new Intel Xeon Phi technology for such task. The source code proved to have many bugs and as such it took a long time to be able to produce consistent results. Thus, the goal of the project was not completely fulfilled. However, the code was ported to Intel Xeon Phi and showed that the programming effort required to use the card is indeed less than that of GPU. However, further optimizations are needed to get a good performance.