Molekulardynamik auf Hochleistungsrechnern : Selbstorganisation und dynamische Strukturen in Flüssigkeiten

Gujt, Jure LSF; Spohr, Eckhard GND

Dieser Artikel erklärt, wie mit Modellen, Hypothesen, Computern und Programmen Forschungsfragen beantwortet und neue gestellt werden (können).

Why does chemistry need large computers? Because the interactions between individual atoms and molecules determine the behavior of chemical systems, and because the numerical calculation of these interactions is difficult and numerically expensive. Computer simulations of these interactions form one of the pillars of modern natural sciences. Computers are particularly useful when it comes to disordered systems such as liquids and solutions in which hierarchical processes of selforganization take place on different time and length scales. A current research project of our group is the study of the formation and fusion of so-called micelles (which are, in 36 turn, self-assembled structures in aqueous solutions composed of many molecules with a hydrophilic (water-loving) and a hydrophobic (water-hating) end) into cylindrical and three-dimensional entities. The dependence of this behavior on small ions with specific activity offers the potential to manipulate the shape and dynamics of the micellar systems themselves. Instrumental and essential for such studies is the use of computer systems of varying computational power, which we demonstrate on the basis of our experience with the performance of server PCs in our group and the supercomputers CRAY-XT6m and NEC-magnitUDE at the Center of Computational Sciences and Simulation at our university.



Citation style:
Gujt, J., Spohr, E., 2017. Molekulardynamik auf Hochleistungsrechnern: Selbstorganisation und dynamische Strukturen in Flüssigkeiten. Natur-, Ingenieur- und Wirtschaftswissenschaften - High-Performance und Cloud Computing.
Could not load citation form.


Use and reproduction:
All rights reserved