Currently, petaflop machines are state-of-the-art parallel computing environments. However, the US Government is already aiming at exaflop devices. Despite the ever growing massive parallelism and speed, the development of codes that can use these resources efficiently are highly nontrivial to develop. Because problems in various disciplines of science require vast amounts of computer time, the development of efficient algorithms that scale on these massive machines is of great importance.
Contacts
Xiaofeng Qian
Joint Appointment in Physics and Astronomy, Associate Professor of Materials Science and Engineering