CPMD 2017 Workshop

Last week, I was able to attend the CPMD2017 workshop in Tsukuba, Japan. CPMD in this case stands for Car-Parrinello molecular dynamics, one of the most important techniques in computational physics. This workshop featured talks from some of the top scientists in the field, including of course both Roberto Car and Michele Parrinello.

This workshop also coincided with the 70th birthday of Dr. Roberto Car. It was fun to see people celebrate as he certainly had a big impact on the scientific careers of many people there. On a less positive note, this workshop all coincided with Typhoon Lan. I hope that everyone from the conference was able to safely and smoothly return home despite the severe weather.

What is CPMD and what it is useful for? First principles techniques such as density functional theory can be used to compute the ground state energy of a given molecular system. This works well for computing things like bulk crystal systems, but what about dynamically evolving systems such as liquid water? In that case, we need to be able to compute the motion of atoms as well. This combination of molecular dynamics and density functional theory is an important tool for interpreting experiment, and studying systems under extreme conditions. Car and Parrinello were pioneers in combining molecular dynamics with density functional theory, including inventing the method now called Car-Parrinello molecular dynamics which was first introduced in their 1985 paper [1].

For me personally, the most impactful talk was given by Dr. Osamu Sugino from the University of Tokyo. In his talk, he referenced his group's work on using first-principles molecular dynamics simulations to study firefly bioluminescence [2]. For this study, it was necessary for them to relax the structure of the oxyluciferin anions in solution. This simulation proved to be very slow to converge, but in a true tour de force calculation they spent 200 days performing this simulation using 128 Xeon processors for each isomer. These time consuming simulations are an inspiration to us who work on the computational side, reminding us how important it is to get the best performance out of our codes.

[1] Car, Richard, and Mark Parrinello. "Unified approach for molecular dynamics and density-functional theory." Physical review letters 55, no. 22 (1985): 2471.

[2] Noguchi, Yoshifumi, Miyabi Hiyama, Motoyuki Shiga, Osamu Sugino, and Hidefumi Akiyama. "Reverse Stability of Oxyluciferin Isomers in Aqueous Solutions." The Journal of Physical Chemistry B 120, no. 34 (2016): 8776-8783.