Atomic Works & Professor Ian Hamilton are developing software that combines quantum and classical mechanics

Atomic Works and Professor Ian Hamilton from Wilfrid Laurier University are teaming up for a second time to develop models that combine quantum and classical mechanical theory to optimize the discovery and development of new molecular compounds.

 

Atomic Works is a Mississauga-based company whose software employs advanced Computer-Aided Design (CAD) methodologies, machine learning algorithms, and neural networks to predict the best design for a given molecular compound. The company initially teamed up with Professor Ian Hamilton and his research team at Wilfrid Laurier University in 2017 to test the limits of their computational modelling. Their work has expanded into a multi-year collaboration that explores how quantum mechanical and classical mechanical modelling can be efficiently combined at scale.

The findings from Professor Ian Hamilton’s team will allow Atomic Works to develop an improved front-end to the open-source software CP2K (the industry-standard for atomistic simulations of solid-state, liquid, molecular and biological systems). If the collaborators can overcome the hurdle of properly scaling message passing between CP2K and Atomic Works’ front-end software, Atomic Works stands to win big as the go-to software for modelling all kinds of physical structures, across a variety of disciplines.

Their first collaboration, led by SOSCIP TalentEdge Post-Doctoral Fellow, Dr. Christopher Ehlert, successfully improved preprocessing times for message passing between CP2K and Atomic Works’ software. It resulted in a newly commercialized product and service for Atomic Works. Their second collaboration aims to further improve the preprocessing time for CP2K, again with the potential for new commercialization opportunities.

Ultimately, the collaborators aim to completely optimize the time spent between juggling complex calculations and sending the resulting data to Atomic Works’ software suite.

“We aim to explore the most efficient way to break down large-simulation domains,” Hamilton says, “and optimize message passing between computational threads to achieve a near-linear scale increase in performance.”

In the current phase of their collaboration, which will run until 2022, the team has expanded to include Ph.D. students, Heather Gaebler and Sylvia Smith. Through their collaboration with Atomic Works, Professor Hamilton’s students gain experience that is widely applicable both in Academe and  Industry. Regular interaction between the Wilfrid Laurier team and their partners at Atomic Works is translating into improved presentation skills, critical and creative thinking, data management and time management for the students involved.

Moreover, the SOSCIP Parallel-CPU platform provides Professor Hamilton and his team with experience on a large-scale computational system. Having access to such a system is essential to their work, which relies on running 5,000 density functional calculations every two femtoseconds.

“SOSCIP resources have enabled us to do calculations that just wouldn’t have been possible to undertake otherwise.” he says.

The practical implications of Professor Hamilton and Atomic Works’ collaboration are specifically evident in the modelling of so-called ‘quantum dots.’ These species have applications in photonics where they can be used to generate electricity from light (in solar panels), and light from electricity (in display screens).

However, Professor Hamilton also sees significant potential for magnetic quantum dots as quantum bits (qubits) in quantum information processing. With the ability to model magnetic quantum dots efficiently and at scale, Atomic Works will be well-placed to ride the oncoming wave of society’s transition from traditional computing to quantum computing.