Hide menu

TDDC78 Guest Lecture (2013)


Case Studies of Parallelising Real Scientific Applications

Presenters: Soon-Heum "Jeff" Ko and Chandan Basu (National Supercomputing Center, Linköping University)

Tuesday 21 May 2013, 13:15, lecture hall C4

Abstract:

The continuous advancements in computing technology make numerical simulation a powerful strategy for the understanding of complex scientific phenomena. On the other hand, the frequent invention of new computing techniques makes it challenging for scientific users to keep updated. Therefore, NSC has staff working as application experts and computational scientists, who can help domain scientists take advantage of high-end computing technologies. As a main SNIC (Swedish National Infrastructure for Computing) center and a proud PRACE (Partnership for Advanced Computing in Europe) partner, NSC (National Supercomputing Center) performs a number of projects to scale up scientific application codes.
In this talk, we introduce a couple of NSC's activities to apply advanced computing techniques to scientific application codes. This talk consists of three subtopics, each of which will explain individual steps of parallelism:
  • How to prepare the parallelisation from the sequential code: An example of analyzing a hydrological environment code
  • Design and development of a parallel code in effective manner: An example of parallelising a quantum chemistry code
  • Scaling up the performance of parallelized codes: An example of improving a bio-informatics code

Page responsible: Webmaster
Last updated: 2013-05-16