Blog Details

Scientists turn to the cloud to streamline supercomputer calculations for chemistry

A team led by researchers from the Pacific Northwest National Laboratory is finding new ways to accelerate the pace of computational chemistry, by making tools for quantum computing and AI-assisted data analysis available via the cloud.

Their effort to make supercomputer-scale resources more widely available through cloud computing could aid in the search for methods to break down toxic “forever chemicals” that are currently hard to get rid of. And that’s just one example.

The researchers describe their progress on the project — known as Transferring Exascale Computational Chemistry to Cloud Computing Environment and Emerging Hardware Technologies, or TEC4 — in a study published today in the Journal of Chemical Physics.

“This is an entirely new paradigm for scientific computing,” PNNL computational chemist Karol Kowalski, who led the cross-disciplinary effort, said in a news release. “We have shown that it’s possible to bundle software as a service with cloud computing resources. The initial proof of concept shows that cloud computing can provide a menu of options to complement and supplement high-performance computing for solving complex scientific problems.”

The TEC4 project makes use of resources including Berkeley Lab’s Perlmutter supercomputer, Microsoft’s cloud-based Azure Quantum Elements platform and NWChem, a computational chemistry software package developed at PNNL.

Study co-author Nathan Baker, product leader for Azure Quantum Elements, said Microsoft’s collaboration with PNNL “is a great example of how modern AI and HPC [high-performance computing] tools can advance computational chemistry.”

In their proof-of-concept exercise, the researchers ran a simulation of chemical interactions involving perfluorooctanoic acid — which is one of the industrial “forever chemicals” that was historically used for the production of non-stick coatings and firefighting foam. The chemical, also known as PFOA, was banned worldwide in 2019 but still persists in the environment.

The simulation traced a step-by-step process that broke down some of PFOA’s molecular bonds. “This sequence of reactions highlights a potential pathway for the degradation of PFOA, providing valuable insights into the chemical behavior and breakdown of this persistent environmental pollutant,” the researchers reported.

Now the researchers are recruiting additional collaborators to put their cloud-based platform to more rigorous tests. “We are building a family of codes,” Kowalski said. “The goal is to build a community around this effort.”

A graduate-level course that builds on the TEC4 project has already been developed at the University of Texas at El Paso, in collaboration with PNNL and Central Michigan University. Kowalski said TEC4 is pointing the way not only to new chemicals, but also to new ways of doing chemistry.

“We envision an ecosystem of use cases from low-tier to high-tier jobs that take advantage of GPU-based computing now being used extensively for artificial intelligence and machine learning applications,” Kowalski said. “We want to allow users to take advantage of different layers of compute, paying only for what’s needed and bundling software with compute access. This is the first step toward that future state.”

Kowalski and Baker are among 32 authors of “Electronic Structure Simulations in the Cloud Computing Environment,” the study published in the Journal of Chemical Physics. Other authors represent PNNL and Microsoft Quantum as well as Berkeley Lab, Argonne National Lab, Central Michigan University, the University of Texas at El Paso and the University of Washington.