Molecular Dynamics Simulations

MASSIVELY PARALLEL CODE FOR MULTISCALE MOLECULAR DYNAMICS (MD) SIMULATIONS

LEAD PI: DR. PAOLO CARLONI

MENTOR: DR. VIACHESLAV BOLNYKH

This collaborative project aims at the further development of our existing massively parallel code for multiscale molecular dynamics (MD) simulations, MiMiC.^1,2 The candidate will work on the performance optimization (parallelizing sequential algorithms, porting existing ones to GPUs or work on the optimization of existing solutions). 

The expected outcome of this project is a new software module for our open-source package that either implement a novel algorithm/model or improves the performance of our previous solutions. Successful implementations may also be published in a high-impact journal. 

This collaboration will be part of a long-term project connected to the development of the MiMiC code, together with Prof. U. Roethlisberger (EPFL, Lausanne, Switzerland). This project has existed for few years and is going to be supported in the future through multiple initiatives and grants. Therefore, this opens opportunities for further longer collaborations with successful candidates. 

REQUIRED SKILLS

Successful candidates should have:

• Master’s in physics, chemistry or computer science
• Basic knowledge of Fortran or C++
• User-experience with Linux OS
• Some understanding of Python 
• Familiarity with MD computational workflows is a plus

COMPUTATIONAL RESOURCES

During the collaboration, the intern and their mentors will be given access to machines located at Jülich Forschungszentrum and RWTH Aachen University IT Center. This gives a unique opportunity to get a hands-on experience with some of the novel supercomputing architectures such as the modular cluster-booster system available at JSC.

REFERENCES

1. Haugaard Olsen J. M., Bolnykh V., Meloni S., Ippoliti E., Bircher M. P., Carloni P., and Rothlisberger U., MiMiC: A Novel Framework for Multiscale Modeling in Computational Chemistry
   Journal of Chemical Theory and Computation 2019 15 (6), 3810-3823
   DOI: 10.1021/acs.jctc.9b00093
2. Bolnykh V., Haugaard Olsen J. M., Meloni S., Bircher M. P., Ippoliti E., Carloni P., and Rothlisberger U., Extreme Scalability of DFT-Based QM/MM MD Simulations Using MiMiC
   Journal of Chemical Theory and Computation 2019 15 (10), 5601-5613
   DOI: 10.1021/acs.jctc.9b00424