报告题目:Wave Dark Matter Predictions from GPU-accelerated Adaptive Mesh Refinement Simulations

报  告  人:Hsi-Yu Schive, National Center for Super computing Applications, USA

报告时间:2018-01-18 14:00

报告地点:蒙民伟科技南楼 S727

报告摘要:The conventional particle interpretation of cold dark matter (CDM) still lacks laboratory support and struggles to explain the basic properties of dwarf galaxies. This tension motivates wave dark matter (ψDM) composed of extremely light bosons (mψ~10-22 eV), which suppresses structure below the kpc scale by the uncertainty principle but retains the large-scale structure predicted by CDM. In the first part of this talk, I will present the first cosmological ψDM simulations that achieve an unprecedented high resolution capable of resolving dwarf galaxies. These simulations reveal that every ψDM halo has a prominent soliton core surrounded by fluctuating density granules. These predictions compare favorably with the observations of galaxy formation, the Lyman-alpha forest and reionization, and also help explain gravitational lensing flux anomalies. The second part of the talk focuses on GAMER, a GPU-accelerated adaptive mesh refinement (AMR) code. A rich set of physics modules is incorporated and which outperforms other widely-adopted AMR codes by one to two orders of magnitude. The code scales well to thousands of GPUs and achieves a uniform resolution as high as 10,2403 cells. I will present several ongoing astrophysical projects with GAMER that require substantially higher resolution than previously feasible, including turbulence cascade in galaxy cluster mergers, star formation in isolated disk galaxies, supermassive black hole accretion, and ψDM simulations.