Title: Accelerated algebraic functions for statistical genomics
Authors: Alexander Freudenberg - University of Mannheim (Germany) [presenting]
Abstract: Genomic datasets used in empirical research are steadily growing in size and advances in computing power can only partially offset the associated computational demand. A suboptimal utilization of hardware resources can lead to significant increases in costs and computation times. We explore the benefits of highly finetuned GPU routines for the calculation of important population statistics, which utilize instruction sets of modern NVIDIA hardware. We showcase examples as part of our implementation in the R package miraculix.