University of Washington - Department of Statistics
The advent of computer clusters within the department and elsewhere has made available large numbers of CPUs to aid numerically intensive computing. One way to harness this is via software to parallelize computational tasks. However, while straight-forward in principal, the "user's manual" for those wanting to use parallel methods in C or R is hard to find. In this didactic seminar, an expert (Hana Sevcikova), a simplistic user (Mark Handcock) and special guest star (Dustin Lennon) will share their experiences with PVM (Parallel Virtual Machine) and MPI (Message Passing Interface).