Case Study: HPC Cluster Speeds Molecular Dynamics Research

In our latest case study, researchers at the University of Richmond shares details of their use of an HPC cluster built by Advanced Clustering Technologies to expand our understanding the dynamical behavior of interesting molecular systems.

A research group at the university, led by Chemistry Professor Dr. Carol Parish, needed a high performance cluster to support the large number of students working on a diverse range of research projects. Advanced Clustering provided a 22-node cluster including NVIDIA GPUs.

Using this cluster, the team is running molecular dynamics and quantum mechanics calculations.

“One of the pros of having GPUs is that we can run our jobs 10 times faster than on CPUs,” said Erica Modeste, HHMI Post Baccalaureate Research Assistant. “A lot of communication needs to happen between nodes. The cluster Advanced Clustering Technologies built for us has doubled the speed of our timestamp. We could get 15 nanoseconds per day with CPUs. We get 25 nanoseconds per day with GPU, which is a big deal in computer world.”

Read the full case study.