4 p.m. - 5:15 p.m. Location: SLC 1.102
Dr. Salman Habib (Argonne National Laboratory)
The formation of structure in the Universe is one of the key problems in cosmology. Over the last 50 years, computational approaches to this problem have been essential in improving our understanding of the underlying physical processes. In addition, they have produced quantitative results that, while powerful, are not yet fully understood. Cosmological measurements have made great strides over the same period. Large-volume sky surveys have accessed the vast temporal and spatial expanse of the Universe via a remarkable set of observations, and many more are sure to follow. To make new predictions for these observations, and to properly interpret them, large-scale numerical simulation and modeling, combined with new statistical methods, continues to be an essential tool. In this talk, I will review the current situation and describe the development of HACC (Hardware/Hybrid Accelerated Cosmology Code), an extreme-scale N-body cosmology code. Since cosmology is an observational science, all we can know is essentially a problem of inference, typically involving a large number of parameters, the so-called "curse of dimensionality". I will also describe the development of the Cosmic Calibration Framework, designed to overcome this problem.