Linear scaling algorithms: introduction, accuracy, and scalable implementation

16 July 2009 (Thu) 10:30 – 12:00

8F large seminar room, Sengen site

Dr. Vincent Sacksteder

Asia Pacific Center for Theoretical Physics, South Korea

Linear scaling algorithms: introduction, accuracy, and scalable implementation

This talk introduces the main ideas of linear scaling algorithms, which promise to extend first principles DFT calculations to large numbers of atoms and large parallel supercomputers. Physically speaking they are implementations of the idea callednearsightedness. I argue that linear scaling algorithms can be exponentially accurate even in metals as long as there is some disorder. Lastly I discuss the enormous difference between having a scalable algorithm and having an implementation which scales to the largest supercomputers, and propose development of a general purpose library for linear scaling codes.

Dr. Tsuyoshi Miyazaki(宮崎剛)