Project Title: Hydrodynamical simulations of galaxy formation in the Planck cosmogony

点击翻译按钮获取中文摘要

The basic model for how galaxies form within the framework of a cold dark matter cosmogony has been established for many years, however many crucial aspects are still poorly understood. For example, what physical processes determine galaxy stellar masses and galaxy sizes? How do these properties evolve throughout cosmic history? How do stars and AGN regulate the evolution of galaxy properties? Numerical simulations and theoretical models are a valuable tool for exploring these questions, but the huge dynamic range involved, and the complexity of the plausible underlying physics, limits the ab initio predictive power of such calculations. The Eagle simulations1 , performed on the Dirac supercomputer in Durham and on curie in Paris, produced a population of galaxies with properties very simulated to those observed. eagle has been a tremendous improvement over previous simulation projects, however the limited numerical resolution still prevents us from examining important aspects of galaxy formation. For example what are the detailed properties of the high redshift, z 10, population of galaxies that are a major target for the James Webb Space Telescope and are thought to be the driver of reionisation, a key science area of lofar. Can simulations reproduce the detailed properties of Milky Way-like galaxies (bulge/disk properties, occurrence of bars, stellar halos, properties of satellites of galaxies) - an important research area in the context of gaia. Simply increasing resolution is not an option because current codes do not scale to the thousands of cores of current supercomputers, nor can they exploit the new generation of hardware accelerators such as intel's Phi and nvidia's GPU. This project will take advantage of the new swift code, developed in Durham in collaboration between the icc, Computer Science, and intel in the context of Durham's status as Intel Parallel Computing Centre (IPCC). swift2 uses task-based parallelism, asynchronous communication, and is designed to take advantage of AVX, with extensions to Phi and GPUs in progress. You will contribute to developing swift, in particular adapting it to include the galaxy formation modules from eagle. You will then use the new code to perform zoomed simulations of individual galaxies and clusters of galaxies at unprecedented realism. Comparing these to observations will provide crucial tests of the CDM paradigm on the small scales that are not currently accessible to simulators.